Slay The Spire: Completing Ascension 20 on All Characters Summary
Overview
I typically pick up one mobile game at a time, play it a ridiculous amount, stop and move on. Over the years its been SimCity BuildIt, FIFA, Brawl Stars, Gwent, Mini Metro / Roadways, and Into the Breach. My most recent obsession has been Slay the Spire. I was looking for a turn based strategy game, this roguelike deck builder is both challenging, replayable, and rewarding. Well, 253 game hours later I’m ready to end its tenure as my main game and move on (though these daily challenges do look fun). I got each of the four characters to Ascension 20 and beat it (I however didn’t really consider The Corrupted Heart as part of the main game)
It’s crazy to me that there are still 3 relics I haven’t unlocked / discovered yet.
I’d like to make a bit of a disclaimer that I’m also very narrow minded in my strategies. I found some routes that worked for me, and this isn’t necessarily the best guide. Watching some YouTubers, I am not nearly as adventurous or successful as them. My win rates are way lower than them. Regardless, here is my experience with Slay the Spire.
Statistics
Slay the Spire collects a lot of run data via the Statistics section of the game, but there wasn’t a good way to export it, so I kept it very high level.
First off, I did the worst with the Defect, as in, I had the most runs with them at 210 or 43%, but when assessed by hours it’s only 38%; probably because I ended up brute forcing my way up the Ascension ladder. A general theme you’ll continue to hear about, is play fast and if you’re gonna lose, die early.
Both Ironclad and Silent averaged about 1 hour for winning runs (59m and 60m respectively), while Watcher was just over at 65m. Defect was the longest at 81m.
Obviously the game gets more difficult with the higher Ascensions, each Ascension adds one layer of difficulty. A12, A13, and A15 also proved to be more difficult. At A17-19 all enemies become more challenging. A20 is really hard, with two final bosses. At around A17, I found that I had to start reading Reddit and other places for some advice. Having a concise strategy and deck became much more necessary, hard to randomly win at these levels.
This is a table of the number of runs it took me to beat each Ascension by character.
I started the game with Ironclad, so it might be a bit unfair, since I didn’t realize exactly how the game worked.
You can see I really struggled with the Defect in A17 and A20. Clearly, I got pretty lucky in between. Winning Silent’s A20 on the first try felt really good though. And I’ll be honest, I basically cheated on A20 for Watcher and Defect, as I won both with the starting relic Snecko… so that’s a bit unfair.
This chart below was the order in which I played.
Color coded by character type
The same chart, broken out by character
Not worth it's own chart, I beat the final bosses at the following rates:
Awakened One - 68%
Time Eater - 74%
Donu & Deca - 86%
Generally I came to be very excited whenever I saw the Donu & Deca symbol, because it felt like it would be the easiest win. When I lost, I lost to a boss 28% of the time. The Act III bosses are definitely much harder in relation to the rest of their act.
I also collected stats on my wins (though some of the early ones were missing, not sure if they were too long ago, or if they simply don’t track all the data). I averaged 24 - 25 cards per win, but a general theme I saw was as the Ascensions got more difficult I had to get better at the game. Getting more particular with card selections and making the deck more optimal. In the beginning of the game my average was closer to 30, whilst by the end I was in the lower 20s. My winning A20 run average was 20. Watcher had the smallest deck size consistently (which makes sense for the optimal dance stance strategy).
Also in the second chart, you can see that the number of upgraded cards also increased. Upgraded cards were nearly at 50% for Ironclad, while it was 35 - 40% for the other three characters. My guess is that because the Ironclad has automatic healing via the starting relic, so I could spend more rest sites on smithing.
Interestingly, in the third chart, you can see the number of relics I had on winning runs dropped over the course of the game. Averaging 20-21 for the game, it eventually dropped to 15 - 18 for the final few Ascensions. My hypothesis here is that I eventually made the decision that I couldn’t efficiently beat Elites in Act II and III, and it wasn’t worth risking the whole run. There are just too many relics that weren’t that great and worth the tradeoff in health or risk of death.
I was surprised that the Silent held onto their Starting cards way more often than any of the other decks. Similarly, I removed them the most for the Defect and Watcher, as they generally didn’t fit in with the deck types I was trying to build.
Also, as you can see, I would lose most often in Act I. Generally by Act III I would have a decent designed deck and strategy. Because my goal was to simply Ascend (not winning streaks), I generally brute forced my way, and if I was going to lose, I figured I should lose fast and early.
Qualms
My Nemeses, these guys would just constantly ruin my runs.
My Weakest Character
The Defect is statistically my worst, but I actually quite liked the character, because when you got its engine going, it was a lot of fun.
I ended up disliking the Watcher the most because by the end of the game it felt like I had to just stance dance as efficiently as possible to win. Wasn’t exactly a lot of fun.
Singular methods…
Maybe I’m just not good at this game, and brute forcing my way to the end wasn’t conducive to being more flexible, but it became frustrating just trying over and over again to try and get the right cards (and relics) to win. Sometimes it didn’t feel as much as my success as just random chance.
Which is kind of why I like Snecko so much. Getting this relic right off the bat, opened up the game, and every turn and every combat felt like there was a chance. I spent much longer on Snecko runs because I truly was considering my actions and being much more careful.
I imagine if I start doing the daily challenges, I will also learn some new strategies and synergies and get better.
Scoring
Statistically I’ve ignored the scores, I don’t really think its a very accurate rating of how strong the run was. Reading about some of the methodology just seems insane.
A Turn Counter
Would be helpful and amusing to see how long certain encounters go on for.
A Rewind / Replay
Sometimes I miss what has happened, or don’t totally understand it. A log / order of operations / description to check would be helpful to better understand the game mechanics.r
Just an absolutely dominating run, this is what invincibility looks like.
Things I Wish I Knew When I Started
Elites
Yes, you should fight them as much as possible, but at higher ascensions (and with more experience under your belt) you’ll know if you have a decent shot or not at beating the boss. So skipping elites in Act III became a common strategy. Earning a random (most likely) less than helpful relic isn’t worth the risk.
Relics
Goes hand in hand with above, but many of them aren’t really that helpful with your specific deck. That said, in later Ascensions, it really pays off to pay attention to anything with a counter. Timing these bonuses can be the difference between winning and losing. In early runs, I just never paid attention, collected them to fill the screen, and just let positive random things happen. Doesn’t work that well in later Ascensions.
Rests
Don’t rest in Act I. This means being more defensive than you may like. But this gives you the opportunity to upgrade some cards that will pay off over the long run. Plus, if you’re brute forcing like I was, resting can prolong a mediocre run. Might as well see if you can just earn a lot of the health back with the rest after the first boss.
Boss Rests - % recovered
I’m not sure why it took me awhile to figure this out, but after Ascension 5, you only heal 75%. So you want to end each Act with minimal health to reap the most benefit.
Drawing Cards
Early on I focused on getting lots of energy and high energy cards. I learned later that having a decent card draw to make sure you always get good cards was way more valuable. What’s the point of having a lot of energy if you don’t have the cards to spend?
Use Potions
Can’t take them with you when you die. My rule was to try and use them if I ever had no potion slots left. Post Ascension 11, you get fewer slots, and you usually save one for a great potion for a boss fight, but try and burn them earlier rather than later.
Artifacts
I never really understood the value of the Artifact. In random early play it would just block a negative effect from an enemy. I never cared for it. It wasn’t until accidentally stumbling upon using it in combination with a card that had a negative effect that I really understood its value in (Defect’s Core Surge combined with Biased Cognition was a great way to fast boost the orbs).
Plan for the Final Boss
Once you start Act III, look at the top and know who your boss is. At some point this is vital. Donu & Deca usually meant I could skip through the Act avoiding elites, and knew I had a decent chance. Time Eater’s card playing limit and Awakened One’s Curiosity Power are huge variables that will actively change the way you want to approach those encounters.
Card combos
Every time I watch a YouTuber show a run, I’m always surprised by some of their card combinations. Synergies between cards that will make both even better.
I also like getting a bunch of the same card, and pushing the overload to “break” the game.
Game Changing Cards / Relics
Snecko is fantastic. It opens up the world to all the high value cards, and makes each combat and turn way more interesting. I only usually was excited to have it, if it showed up in the early game, otherwise it would be too late to effectively pivot your deck.
Ice Cream is delightful both in game and in real life. Being able to save your power from turn to turn just makes your run so much more efficient.
Apotheosis is the only colourless card I would take. It upgrades all your cards in your deck. Huge value if you can get it out early.
Stacking some crazy poison damage.
My Forever Advice
You can’t control really which relics you get. You can mostly control what cards you put in your deck. Manage and optimize your deck.
Remove Cards
Constantly remove cards when given the chance. Valuable cards showing up more often is critical in long fights. At the very least you can do better than the starting cards. I like to think of my deck count also in terms of how many are there in a steady / mature state. This is size of the deck, post Exhausts and Powers. I like to have it around 10 because it means you’ll see every card at minimum every two turns (or fewer if you have some draw).
Don’t Add Cards
If a card doesn’t substantially improve your deck, then don’t take it. I often consider if this card actively synergizes with my current deck and if it is something I want to have in my deck in Act III.
As my progress stalled, I often referred to Reddit for Relic / Card Tier Lists. So here is my Relic Tier List:
S, A, & B are great, C are fine, D and F which are cropped were useless to me normally.
Strategies
Ironclad
Strength is the name of the game for Ironclad. It’s straightforward. Get lots of strength and then obliterate your enemies. Demon Form and Limit Break are great ways to get your strength boosted. On the reverse end you want to invest in blocks so you can survive as this strength builds for big fights. Just make sure to have a Heavy Blade so you can fully take advantage of that attained Strength.
The only other deck I had some success with was Perfected Strike. If I started getting those cards early, I ran with it, and just only drafted cards that had Strike in it. They’re super strong in the early game.
Top Five Cards:
Silent
This is the easiest character; there are really only two types of decks for this character; Poison or Shiv. Depending on the cards, I usually just start going hard to one side or the other. Generally I preferred Shiv (10/16)
Poison relied heaviest on Catalyst, which would double or (if upgraded) triple the poison. Corpse Explosion is also a must, to help you transfer over all that hard earned poison onto other enemies. It feels pretty good when you can get Poison over 1,000.
To go Shiv deck, you really needed Accuracy cards (+4 or +6 for shivs), but any relics or cards that were focused on rewarding lots of attacks / cards being played would end up being very good. Timing a big attack turn with Phantasmal Killer was clutch.
Top Five Cards (tough, since there are two explicit routes you can take, so I’ll do one for Poison and one for Shiv) (also special shoutout to Nightmare, which if timed, can be devastating):
Defect
This character was tough, in earlier ascensions, I seemed to win with all sorts of strategies. By the end there was really only one… Frost and Echo Form. Get lots of frost going, and build up a constant flow of a lot of defence, then slowly chip away at the enemies HP. Getting a lot of focus and orb slots really helps with this. Orbs max at 10, so don’t bother trying to get more than that. It feels really good when your Frost engine is churning so much defense that you become invincible.
Watcher
Similar to Defect, by the higher ascensions, I had to get much more consistent with my strategy. In this case, it was to stance dance and create highly infinite decks. The trick is to get the cards that make sure you get energy every time you leave Calm and to draw cards every time you enter Wrath. And if you’re careful you can keep this rolling for many many turns. Keep refreshing energy and card draw with a small deck, means you can really control the tempo. Mental Fortress is clutch, as well as Fear No Evil and Inner Peace are great for exiting Wrath. And Tantrum was a great way to get an enter-Wrath card as conveniently as possible. I really liked Cut Through Fate and would always take it. It helped with card draw, extending out current turns, and planning out future turns. Plus it still had a standard base attack.
Earlier on, I got to successfully win with Alpha, Smite, and Divinity decks, but they really weren’t very successful at the highest ascensions.
Top Five Cards:
Overview
This is an entertaining and challenging game. Remarkably replayable and balanced. Each of the characters plays so differently, and winning feels really good. At points trying to get this far with each character has been a slog, but overall it has been an enjoyable ride.
The Value of a Regular Season Game
Background
It’s playoff season autumn here in New York City. The Barclays Center in Brooklyn is hosting the New York Liberty against the Minnesota Lynx in the WNBA Final. Queens has the Mets at CitiField against the Los Angeles Dodgers in the NLCS. The Bronx has the Yankees against the Cleveland Guardians. Hudson County (New Jersey) has Gotham into the first round of the playoffs next week. Even MLS’ New York City FC and New York Red Bulls are both likely to qualify for the playoffs as well. Clinched at their current standing of 5th and 7th with a game remaining.
It’s obvious to anyone that ticket prices for the end of these team’s seasons are much higher than earlier in the season. I paid $18 a ticket (with fees) for the Mets to play the Dodgers (for a double header no less) in May (and we could have gone for even less. The same matchup in October in the playoffs is minimum $250. Even within the playoffs, we went to the semi-final round of the WNBA playoffs for $37 a ticket (with fees) and the finals were $75 each.
In my opinion the ideal situation is to get a high importance game for the best price. It means the stadium energy will be higher and make the attendance in person more fun than a “standard regular season” game. Sometimes that’s a home opener, a rivalry game, a theme night, or a highly consequential match. This final category made me wonder how teams and leagues could get more high stakes matches. The obvious examples are the US Open (soccer or tennis) or March Madness (though this could be argued it’s just a playoff of the broader NCAA system). Single elimination tournaments where very game matters. You lose, it’s over, try again next year. However while entertaining, these systems aren’t the most sustainable for businesses. Revenue would vary wildly year to year if this was their only source. So, I understand the practical necessity of group stages or regular seasons when designing a tournament or a league.
I began to contemplate league formats. Something that we take for granted, has been proven in recent years, to be entirely flexible. On the positive side, we have the new NBA Cup, which is a fun in season tournament based on similar soccer style tournaments. It’s a great addition that has a single elimination tournament that really shows the talent and skill of the stars. Lebron James really excels at this in his older age. As his peak athleticism has begun to fade in his later years, he cannot keep up with the long regular season or the long best of 7 series in the playoffs / finals. But in both the 2024 Olympics and 2023 NBA Cup, he showed that he is still an absolute dominant force. There is no managing minutes (I mean technically there still is, but let me have this), it’s winner takes all, and an elite athlete like him gets to go for it.
My general concern though, is a gravitation towards expanding tournaments and leagues by adding games, thereby diluting the value of all of the games. The UEFA Champions League did this by adding two group stage games and American Leagues by adding in more playoff spots with wildcard rounds. The NFL added another regular season game in 2021 (albeit by removing a preseason game to compensate) and NCAA College Football expanded its playoff system to 12 teams in 2024, after expanding it to 4 teams in 2014. NCAA March Madness added the First Four in 2011 and the NBA followed in suit in 2020 adding in a play-in round as well (interestingly branded as play-in, not playoffs, trying to retain some of the original value).
It got me thinking, how could I rank the coefficient of each league’s regular season games? If so many teams qualify for the playoffs, doesn’t that dilute the value of each regular season game? How could we compare a regular season game from different leagues against one another?
Methodology
It’s hard to figure out how to come up with a single number for each of these leagues, so I enlisted the help of my favorite econ professor. After a lengthy debate over what’s the objective of sports clubs, we decided to look at it from the perspective of winning the league to simplify the math (certainly not all teams are actually looking to do this). I wanted to have a singular value for a league’s regular season game. Something I could compare across sports and leagues. We thought we would compare soccer leagues against one another, and then look at the American systems across different sports.
We ignored league specifics, such as how scheduling works. Playing against divisional opponents more often, or higher rated teams / conferences, as it just threw a wrench into everything. We also ignored the number of games in a playoff series. This generally sufficiently provides enough excitement to the game that I’ll let it slide. I’m not a fan, but it’s fine for these purposes.
Yes, higher seeding is marginally helpful, but if a team that is ranked the lowest in the playoffs has the same real chance to win the league, then isn’t the goal just to qualify? Also in the reverse, there are some teams that are basically guaranteed to qualify, given their relative strength to the field. The win % was set to a standard 50/50, a huge assumption that every team is equal and the odds wouldn’t change as the season went on. I didn’t feel like learning what a Markov Chain is and how to implement it. So we stuck with a simpler version that would hopefully provide some insight.
Binomial Distribution is a method of evaluating probability distribution. In my basic view it helps show us what are the odds of getting a certain result over a certain number of independent trials. In relation to this sports question, it would show us how likely it would be to win a certain number of games (to win a league or qualify for the playoffs) if given a certain number of games (the total regular season game count).
Also at any point please tell me if what we’re doing doesn’t make any sense. I barely understand how we got here. Part of the reason for this Methodology section is a means to double check our work. Here are our steps (and I’ll use the NFL as our example in the bullet points):
Collected and organized data on each league, most importantly number of regular season games, number of wins on average required to qualify for the playoffs / to win the league, and if applicable, the number of rounds in the playoffs.
17 Regular Season Games
11 Wins on Average to Qualify for the Playoffs
4 Rounds of Playoffs (if you’re the low seed) or 3 Rounds of Playoffs (if you’re the high seed)
Working our way backwards through the playoff systems (and we assumed 50/50 odds for all matches for simplicity, though we’ll adjust that in some later iterations). We ended up with an Weighted Average Chance of Winning the Playoffs value. (It had to be weighted because most leagues had different round entry points for different teams)
7.14% Chance of Winning the Playoffs (if even) (broken down to 6.25% for the 12 teams who have to do 4 rounds and 12.50% for the 2 teams who have a bye and only do 3 rounds)
Using that Average Number of Wins to Qualify as our “Trial Successes”, we used the formula: =BINOM.DIST((# of Regular Season Games) - (# of Regular Season Wins to Qualify),(# of Regular Season Games),(Odds of 0.5),(TRUE)). Due to the limitations of the function, we had to use TRUE, and flip the goal from Wins to Losses. The True focused our coin flips to return the percentage chance it would take to get fewer or equal to our Successes. So we couldn’t actually do it via Wins, so we swapped it to maximum number of Losses, and it seemed to work fine.
=BINOM.DIST((17 - 11),17,0.5,TRUE)
So now we have the Odds of Qualifying with Random Chance at the Beginning of the Season.
7.17% Chance of Qualifying for the Playoffs at the Begining of the Season (if all games are equal odds)
Combining this with the Odds of Winning the Playoffs, we could then have the Odds of Winning the League
0.51% Chance of Winning the NFL at the start of the season (if all games and teams are equal odds)
However, this doesn’t quite show us what the value of a regular season game is. So we changed the experiment to see what the value of a win would do to the odds.
=BINOM.DIST((17 - 11),(17 - 1),0.5,TRUE) removing one overall game from the season, but not changing the successes, as we’re focused on the number of losses technically.
The first win is changes the odds of qualifying for the playoffs to 10.51% (from 7.17%).
And subtracting those two values, the first win improved the team’s odds by 3.33%.
Now we thought this would be a good way to assess the wins, but after some playing around we found that this was only the correct value of the very first win in the first game. It is not the value of any regular season win. What is missing is if you had a loss. This is where it got more complicated and we sorta phoned it in. We ended up just repeating this method by each win. It’s not quite the answer, but it revealed some interesting charts.
The second win in the second game increased odds to 15.09% and was worth 4.58% and the third win in the third game increased odds to 21.20% and was worth 6.11%.
This would generally scale in every league until you basically guaranteed qualification.
Taking a step back, we thought, there were two simpler metrics to also have on hand that would be helpful; what percentage of teams qualify for the playoffs and what percentage of wins are needed to qualify for the playoffs
43.75% of teams qualify for the playoffs (14 / 32) and a team generally needs to win 65.19% of their regular season games (~11 / 17).
For some other leagues, we expanded this beyond just trying to win the league or get into the playoffs, and added in goals for qualification to tournaments (similar to playoffs) or avoiding relegation. Led to some interesting results, where we even started messing with the odds of winning games.
There are a few huge issues in this:
Primarily, what’s the point of sports and a regular season? It became very apparent that it’s not everyone’s goal or likelihood of winning the season. So if you’re in attendance / supporting a team, then you’re not often not spending your time hoping to win the season. So the likelihood of winning the season doesn’t actually indicate value of attending a game. Arguably a better metric, to answer my original question of which sport I should be attending, is… fun? But we can try and measure that in another post. I’m thinking crowd size, percentage of capacity filled, distance to stadium, cost of tickets, rarity, weather, and other metrics all have factors in “my personal” enjoyment of the game.
Our math doesn’t count wins in relation to losses, as I don’t know what that formula would look like, nor did I care to learn as this had dragged out much longer than I thought it would.
Our variables are all independent, focused on a singular team. Their odds of winning a game and getting enough wins to get past the objective number of wins. But obviously sports are not played in a vaccuum by one team.
Our odds were always set at 50/50 for every match. This just isn’t true in real life, even if you’re about a middle of the pack team, there would be many match ups where you’d be the favorite, and plenty of others where you would be the underdog. To adjust this by sport would take a much larger amount of data, that again, I didn’t have time or the heart to tackle. I did give it a go for the EPL, the league I watch the most, to demonstrate those differences. Really shed more light on things.
Results (America)
Alright, so here we go! First up is a high level scatter plot of the major American sports leagues’ regular seasons. This is a basic comparison showing the percentage of teams that qualify and the percentage of wins in the regular season to qualify for the post season.
The NBA / WNBA and NCAAWB / NCAAMB are nearly identical to one another in structure.
Having more than half your teams qualify for the post season is a bit flawed. Looking at you professional basketball. The NBA tries to differentiate its first round of the playoffs by calling it the NBA Play-In Tournament or a preliminary postseason tournament… so its the first round of the postseason. Cool. In doing so, you dilute the value of your regular season, by allowing two-thirds of your teams into the post season. The major outliers here on the other end of the spectrum are the collegiate sports. This is heavily due to their “non-professional” status (highly debatable), but at the very least the majority of schools that participate exist outside their athletic programs and as young amateurs, they play significantly fewer games than their older professional counterparts. This combination leads to more of their games mattering and fewer teams qualifying for the postseason. But in their infinite drive for more money, in NCAA Men’s Football’s case, this season will be the first season there will be a 12 team playoff (for a decade it was only 4 teams and prior to that just a 2 team final). Adding in multiple rounds of games that they are hoping to attract even larger audiences for.
This is the same chart as above, just zoomed in more. NCAAMF is the top left one, the label disappeared.
Another way to look at this is to subtract the Percentage of Wins Required to Qualify against the Percentage of Teams That Qualify (you want a high percentage of wins and a low percentage of teams to make your regular season most competitive). Four leagues were negative: NBA, MLS, WNBA, and NWSL. Not a great showing for American soccer or Women’s Sports (please save your backlash, I attend in person to mostly soccer and women’s sports games).
The NWSL and MLS have an embarrassingly low threshold for Wins Required to get to the post season, effectively making the regular season pointless. One could argue that soccer’s typical league set up is not conducive to playoffs, but I think it’s more of a conscious choice issue. Never should more than half your teams go to the playoffs. They do have a prize for the top regular season standing team (mimicking most international soccer leagues), but there’s not enough value between that 1st place team and the 18th placed team to justify trying to play hard and win.
Both women’s professional leagues struggle with trying to have full scale postseasons (which have increased viewers and attendances), but don’t have enough teams to justify it, thereby devaluing the rest of the season.
Honestly in hindsight I probably could have just stopped here. This basically sums up my whole point without getting too complicated.
But at the very least I wanted to be able to somehow compare international soccer leagues (without playoffs) to American sports leagues (with playoffs). This is where I needed to set a goal that is universal throughout. For simplicity I went with “to win the league”. Now there was a lot of debate here. It is certainly not every team’s goal to do this and it’s not every person’s viewing reason either. And all teams are not created fairly, but we have to make reasonable standardized assumptions across the leagues to be able to compare.
First I separated the playoffs from the regular season. Odds of winning the playoffs were easy. Take the odds (0.5) and put that to the power of however many rounds (ignoring series / number of games within a series) there were. There was a slight wrinkle that many of these playoffs had play-in / wildcard rounds or gave top teams a bye or free round. Name it whatever you like, they’re basically all the same function. Either way, you can weight the probability by the number of teams, and come up with a weighted average chance of winning (far right column). Obviously the fewer rounds you have and fewer teams you have a higher chance of winning. This is not necessarily a bad thing, if you have a high bar of entry to get into the post season (unfortunately that is not the case for women’s sports).
Squarespace sucks at showing tables, so unfortunately they have to be images (which I hate).
College Basketball (at 50/50 odds) is just absolute mayhem. Probably why it’s got incredibly high viewership. Single elimination tournament? Beautiful in its simplistic competitive design.
The next step was to evaluate the regular season. What are the average number of wins to get into the post season? I took the last 5ish seasons of data (I had to omit COVID related shortened seasons) and came up with the average wins it took to qualify.
A playoff team needs to do incredibly well in the collegiate tournaments, they have few games and need a high winning percentage. Professionally, the NFL has the most valuable regular season games and soccer leagues the least. The MLB requires the most wins, unsurprisingly, since they have the largest number of regular season games. If we just rank these two metrics and averaged it, we can see this combined:
NCAA Men’s Football
NFL
NCAA Women’s Basketball & Men’s Basketball
(women’s is technically a little higher, but only because their average number of wins to qualify is higher, likely because the field is more uneven, so qualifying teams have simply won more games)
WNBA / NWSL / MLS
NBA / NHL
MLB
Those 5th placed leagues get particularly punished for just having a lot of games. If you change the weighting to favor percentage of wins required over number, the NBA / NHL move above the WNBA / NWSL / MLS. The MLB sits again at the bottom…
So this is where that aforementioned Binomial Distribution comes into play. Basically, what are your chances of winning enough games to qualify for the playoffs? Your odds of winning each game here are vital, but for ease of math and in honor of the American Socialism that plagues their professional leagues, we’ll go with fair 50/50 odds for each game.
So league design wise, we can see two huge outliers that are problematic. American soccer… I mean, basically it just shouts, why bother even playing. It’s basically random if you get into the playoffs. On the reverse, NCAA is just ridiculously competitive. There are a ton of schools and you have to do very well to rise above the rest. Now remember, this is at 50/50 odds, and we know that Alabama has better odds than Massachusetts has at football. So it isn’t exactly realistic. Don’t get me started on the insanity that is NCAAMF scheduling. The MLB comes off well here, but it’s simply the slog of 50/50 odds to try and get 95 wins over 162 games. You simply have to be good with better odds to have a realistic chance of doing it.
Alright, here we go, the real meat and potatoes of it all. Combining that odds to qualify for the playoffs with odds to win the playoffs, and you get a team’s chance to win the league at 50/50 odds at the beginning of the season. Pretty unsurprising results. But we can now compare this to the international soccer leagues.
Results (Overall)
Sorry for the false start, here’s the real meat and potatoes of it all. Given random 50/50 odds, at the start of the season a team has these chances to win the league. International soccer leagues were measured based on the number of points (divided by 3 for wins) it would take to win the league (remember there are no playoffs in these leagues).
Now the darker gray bars represent the odds of winning the requisite number of matches to win the league, whilst the lighter grey bar represents just the fact that one of these teams has to actually win. This difference is mainly attributable to the fact that we’ve made the champion truly random at a 50% win rate. It means that it probably would take more teams than are available to succeed at winning if you kept just testing it. We know in reality, that there are plenty of teams that have better odds than 50/50, and their odds will be much better than this. (This makes me want to see what teams had the absolute worst odds and still won the league)
This, however, still doesn’t really answer how “valuable” a regular season game is. So with an econ professor’s vital help, doing this calculation again, but with a free success, we would be able to see the change in chance. However, this change in chance was limited to the first win in the first game, and when repeated, a two game win streak in the first two games, three wins in the first three games, and etc. etc. This is because I don’t know the fancy math to flexibly evaluate all the situations. But at the very least, it should still get us a comparable metric that we could use universally.
I’ll show a quick Sort of the Estimated Average Impact of a Win Towards Playoff Qualification for just the American leagues. I’m going to use the phrase Sort of the Estimated, because we know that this isn’t actually correct, but it’s the best I’m going to be able to do with my current level of intelligence and math acumen. And I want to show this particular chart, before I neuter the American sports by reducing them by their random playoff chances.
Each win improves a teams odds of qualifying as such. Remember this just in relation to their first games. The real outlier is NCAAMF (that light blue line that just rockets upwards), where you basically need to win all of your games to qualify. The fastest declining lines are the soccer leagues. Again, when you start with such high random odds of qualifying, it doesn’t take many wins to basically guarantee you’ll be in the playoffs. Otherwise, typically every league has a nice arc that shows the most impactful games improving your odds are somewhere in the middle of your regular season, where the leagues start to really show separation, are you in the hunt or not. MLB and NHL go past the right bounds of the chart, but they just continue to arc like the others.
So since using just one or the first win as the key metric to compare the leagues by was a bit unfair, we had to blend the league’s season together. See below:
Now taking an average of those, we can sort of estimate the impact of a win. So if the regular season’s purpose is to get into the postseason, and each win gets you a step there, then this is what they’re worth. I showed both average and median, because in particular it was important for NCAAMF since they have so few games, the final wins that get you over the line were hugely valuable and were largely ignored in the median version. Really there ends up being three bands of leagues. Football, the MLB/NBA/NHL, and the rest.
Alright, now let’s show these same two charts, but adjusted for the fact that this only gets you into the playoffs and then you have random odds there. So every league’s regular season value gets capped by the size (odds) of their playoffs.
Large playoffs really hurt the NCAA Basketball leagues the most. The key variable for the rest of these is how few games there are. Football and Women’s sports have fewer and therefore each game is just worth more… feels like I didn’t need to do any of these to come to that conclusion. But, we’re finally getting to the real promised meat and potatoes. Comparing these with international soccer leagues!
Voila. Sort of the Estimated Average Impact of a Regular Season Win Towards Winning the League. As predicted, playoffs make regular seasons much less valuable. A long winded way to prove nothing anyone didn’t already know. Great.
Let’s readdress some of those big caveats we put aside. Not every team wants to win the league. And clearly leagues care about selling more rather than being the best. It may be engrained in our culture, but tuning in for the postseasons where advertisers are paying the most and players are playing the hardest is just how it is. Clearly the broadcasting revenue math is in the favor of playoff systems.
And none of these leagues are actually fair. Despite the socialist safeguards put in place, the odds going into each season and each game are not even. So depending on who you follow and who you’re watching, these odds and impacts may be vastly different. Best to find alternative reasons to watch, than just the path to winning the championship.
The professional leagues in their closed cartel state with pretty evenly shared revenue, losing is still a pretty good way to make money. So unless money truly follows in-game success, then we won’t see leagues optimally designed as such. The not so secret quiet truth about many of these teams in these leagues is that they don’t need to win the championship to be profitable.
Summary By League:
NFL
✩✩✩✩ Rating 4/5
Well if you’re going to award a prize to one of these professional American leagues. The NFL probably takes the trophy. The stats speak for itself, one of the fewest percentage of teams that qualify for the postseason and the shortest regular season of the professional leagues. Add to that that their postseason are single games and not series, and are single elimination tournaments, it just leads to the most excitement, chance, and impact per win. I also appreciate the consistent and predictable scheduling format season after season, building divisional rivalries while still maintaining a rotation of other divisions and conferences.
My improvement suggestion:
would be to cut that first (wildcard) round of the playoffs and just do divisional winners progressing (so only 8 teams qualify for the postseason, although this can be unfair to stronger divisions), or limiting it to the top 4 across their entire conference (this would retain some value for the divisional rivalry games, but be more equitable to teams in strong divisions).
MLB
✩ Rating 1/5
Look, there are just too many games. Its saving graces are that it has the fewest percentage of teams qualify for the postseason out of any professional league and it requires a team to be good to do so. But again… too many games. The average impact of a regular season game was the absolute lowest, by far.
My improvement suggestion.
Just cut some games. It would only help. That said, people seem to love how cheap baseball games are to go to and there are plenty to go to. But, I’ve only been to one playoff MLB game before, and let me tell you, it was infinitely more exciting.
NBA
✩✩✩ Rating 3/5
I mean… it’s fine. The most exciting thing is that just last season they introduced the NBA Cup. Recognizing that their regular season slog was losing interest and wanting to maintain capitalist goals of continually adding games, they threw in the NBA Cup, designed like a soccer tournament. Round robin groups followed by a single game / elimination tournament. My qualm is that they didn’t cut any regular season games to include it.
My improvement suggestion:
Cut the play-in tournament for the post season. Too many teams currently qualify. And probably they should cut some regular season games too.
NHL
✩✩✩ Rating 3/5
Again, it’s fine.
My improvement suggestion:
Add an NHL Cup in season tournament, I think that would be cool. Otherwise, same rules apply. Cut number of teams in the post season and reduce regular season games.
MLS
Rating 0/5
I mean where to begin. Everything is wrong with this league. Way too many teams qualify and you barely need to really win that often to get into the postseason. The Odds of Qualifying for the Postseason at the Start of the Season are nearly 50%, which is inherently pointless. The Leagues Cup was randomly added to showcase Messi a bit more, but it’s kinda crazy since it just interrupts the season and just duplicates efforts of the CONCACAF Champions League without barriers of entry. Complaining about fixture congestion and dropping out of the US Open is embarrassing. The major problem though is that they simply aren’t the best soccer league in the world and at this rate, they’ll never be.
My improvement suggestion:
Do something, anything to improve it. Stop thinking that adding games is what people want. Make games and competitions that matter. Stop trying to be the NFL and stop trying to be the EPL. You’re neither and never going to be either. So be something else.
NCAAMF
✩✩✩✩ Rating 4/5
The winning American league for the value of regular season games. These games matter. You have to basically win them all, and prior to this season, only the top 4 of 134 teams would get into the postseason. Even going perfect (ahem ACC’s Florida State last season) was not a guaranty. We’ll see how a 12 team playoff goes this season, and there will inevitably be changes in the future too. Curious to see how this league progresses (and continually gets corrupted by money).
My improvement suggestion:
Make scheduling and conferences fairer. It’s just the wild west, the NCAA barely regulates this sport anymore. Someone (not a private equity firm) needs to take control.
NCAAMB & NCAAWB
✩✩✩✩ Rating 4/5
Combining these two, since they basically function the same and scored the same barring a slight difference in number of teams and average wins needed. March Madness is arguably the best postseason tournament. Add to that single elimination conference tournaments the month prior and you get a double dose. I hope that the conference realignments and chaos that is engulfing football doesn’t destroy the much more financially stable basketball programs. Because of its great postseasons, the regular seasons aren’t super important, but this could be viewed as a feature, not a bug.
My improvement suggestion:
Nothing really. You could cut a few regular season games and help students go to class more, but whatever, I didn’t go to class often and I didn’t have a good excuse. Can’t really ask for much more. Just pray it doesn’t get ruined.
WNBA
✩✩ Rating 2/5
The WNBA really just mimics the NBA (partially since they are part owned by them), but their team count is just low. Rosters are tight and salaries restrictive (for some good financial reasons), but they need more space to let players grow and competition develop. This video highlights some of those problems (specifically that in the last CBA negotiation salaries increased over ~90% and the salary cap only increased less than ~40%). Apparently a bunch of teams just drop a player on its roster to help stay within salary caps. This leads to a lack of rookies and development, and allows European and Asian markets to take players. It doesn’t help that they’re technically not very independent; originally started as a joint venture with the NBA, WNBA owners only own half of their teams, that plus a financial raise in 2022 diluted their stakes, so they own even less than that.
My improvement suggestion:
Add teams. It will solve most of the league design problems, I know there are financial issues, perhaps creating local divisional structures to limit travel, and offer local incentives and cheap tickets to improve stadium experiences could help mitigate that while it grows. They do have three scheduled to join over the next two seasons which is good. Be the affordable alternative experience people can enjoy.
NWSL
Rating 0/5
Read above for the MLS and WNBA. Just emulating a bad league design in the MLS without the star power or team count to justify it. Add to that the fact that European women soccer leagues are growing, and I seem doom written in their futures. I don’t understand how these valuations keep growing, I’m inclined to call it a bubble. The league isn’t competitive; at the start of the season at 50% odds you have a 60% chance of qualifying for the playoffs and a 7% chance of winning the league. That’s the highest (worst) by far.
My improvement suggestion:
The same model that the MLS used isn’t going to work here and the same models that the other American leagues use aren’t going to work here. The top European teams are currently better invested in both from a talent and financial standpoint and will continually crush the best of the NWSL. Their only chance is to take a financial loss in the short - medium term, and build the NWSL as the pinnacle of women’s soccer internationally. But this is unlikely. So they’ll need to be creative, maybe permanently including Canada and Mexico into the league. Promotion and relegation wouldn’t hurt. Maybe shorten the season and do it twice a year. Replace games with a single elimination tournament. There are a lot of hurdles and pain ahead if things don’t change.
EPL (and other European soccer leagues)
✩✩✩✩✩ Rating 5/5
Yes, I’m biased. Yes, there’s no postseason, but the charts point to the answer. Every regular season game here is more valuable than anything the American leagues offer. Add to that, additional incentives (financial and prestige) to other positions other than winning the league, and you have way way more regular season games that matter. I’ll explore this point in another post.
My improvement suggestion:
Like Germany and France, cutting teams would reduce the number of games (and maintain the perfect round robin structure). It would make the gap between the Premier League and the Championship (the next division) greater, but this could be solved with better equity sharing (if you could pry money from the Premier League’s hands).
And again, just formatted a bit differently:
✩✩✩✩✩ 5/5 - English Premier League
✩✩✩✩ 4/5 - NCAA and NFL
✩✩✩ 3/5 - NBA and NHL
✩✩ 2/5 - WNBA
✩ 1/5 - MLB
0/5 - MLS and NWSL
Conclusion
Leagues are consistently adding games to both the regular season and postseason to increase revenue, however, by doing so, they are increasingly flooding an increasingly crowded sports market with more content and reducing the value of their own regular seasons. This combined with rising costs to attend games and cultural shifts in the way we consume content, is a problematic situation for sports leagues.
The solutions I would propose universally across the board are:
Cheaper tickets and concessions
Fill the stadiums and create an in stadium experience that simply cannot be rivalled digitally
Reduce the number of games
Increases the value of the existing games, both attendance and viewership would likely rise and players would be less exhausted and safer.
If you are going to add games, add more tournaments with real trophies
The NBA Cup is a great example of adding games that players care about and people will watch. To be fair this is all just a precursor to…
Be an international soccer league
Regular season games matter, additional final standings get rewards, promotion and relegation, and additional in season tournaments just make for the most competition and value per game.
Be different
I have long standing proposals that the MLS turn into a Co-Ed league or the MLB include promotion and relegation. I should do posts on both of these. But even broader, just try different things. What if the NBA had a regular season that was first 12 teams past a win threshold? Then players rest and hang out, and there are no garbage games for perennial winners or losers. Just a thought.