Commander Gameplay Stats: The Bonus Data! | Magic: the Gathering

[@devi2289]

Who doesn't love statistics especially about games. Glad we can get another episode about this, the first episode definitely didn't cover everything

[@Tyler-bw3fd]

Who? That's Dr who to you 😁

[@caedmonr112]

Petition for an unlisted video going deep into the stats, p value calculations and etc

[@ryanalytic]

100%. I get very diappointed when they say stuff like, "We won't bore you with the p-values and such." It's not boring: it's literally the part of the video I would be *most* interested in! 😂

[@ryanalytic]

And also because he said colorless decks winning 4/6 games is a "great win percentage". I can guarantee that is not even remotely close to a good sample size to determine if colorless was actually good or if it was just lucky. 😜

[@caedmonr112]

@@ryanalytic oh yeah. That was heavily discussed in the video though (the fact that this study is not nearly large enough for it to be a representable sample, given how much variance commander games have). I'm simply a math nerd who has taken multiple courses on hypothesis testing and I'm interested to see how it would be applied to this fun study

[@ryanalytic]

@@caedmonr112 Yeah, I work in analytics, so my job involves ensuring people understand and can use statistics properly for decision-making. I was just disappointed that, in the same video, he made comments which made it seem like he was making decisions based on a 6-game sample size for certain commanders while at the same time telling people that 35ish games wasn't near enough to tell whether fast mana was good or not in the games he played in. It just gave a mixed message, that's all.

[@ryanalytic]

@@caedmonr112 I also am not going to put in the effort to run the stats on this because I'm not confident I would benefit from such small sample sizes, so who am I to judge? 😜

[@Hamond1989]

As someone who likes working with statistics this was such a fun video, ty :) also Joey being so insecure about people misinterpreting data was kinda adorable

[@daguynexdoor]

I love these stats. If it's not too much work, I'd love to see these stats again next year as well!

[@damonlouis6536]

i need the stats on how often you folded to grave hate

[@chaffenbake]

Thanks for putting your spreadsheet in the description, Joey! Makes it easier to make a template for my own documentation of games!

[@idog8x]

The internet is a weird place Joey, just keep doing your thang ☺️. Thanks for the data!

[@GellyGelbertson]

thanks for the additional stats! i appreciate the continued vigilance against how strongly to consider fast mana, but i still find the results interesting, small sample size or not. also you would be shocked how many of us would want to see you run a chi-squared test against your data - i know i do!

[@neightwards]

Great video Joey! It has been incredibly fun to hear these reviews. I’m also never going to complain when there’s basically a second episode of the EDHRECast in a week (even if the second is a one man show haha). I’m looking forward to hearing this again in 2025!

[@olivierdebilde4292]

Thanks for putting all the time and good work, Joey.

[@crawdaddy2004]

As someone who used to play Legacy, I kinda burnt myself out on playing super competitively. So my favorite decks are the ones that do mostly really dumb things. 😂

[@aeorling]

Magic deck of Theseus! Great info, thank you, Joey!

[@davidhansen5067]

I love stats episodes like this, the data is great food for thought! Please don't become discouraged by those who disregard your disclaimers -- keep sharing!

[@epicwhimsy389]

Already taken your original spreadsheet and adapted it to track my games for this year and beyond. As tend play with same group of 15+ player already expanded to include their win/loss % and same for the commanders. Going to add colour break soon after you included the info on this fab follow up video.

[@jasonritner9662]

I'm curious about what happens when we all start aggregating data like this together from everyone's sources. I started tracking my games this year, mostly to just nerd out about the statistics, but if my information could be fed into someone else's data set to help track trends, it may be interesting to look at that data set for a few hundred people. Shouldn't be taken as gospel, especially if this is all self reported stuff, but it would be fascinating to look at. (I'm tracking the dates of games, what opponents and I are playing, game length, win-con, number of players, "major" deck overhauls, and anything else notable that may have happened, just for funsies)

[@RazgrizAce67]

I'm thinking about trying to start a Discord channel for all those folks like us who log our games to start putting our data together. Would you be interested in that?

[@darkfallen]

I appreciate everything about this process and salute y’all for putting in the effort.

[@lordderyk]

I really loved all of your data. If I had to summarize the entirety of your data, I would say that you had a fun time collecting data.😊

[@jasonhart543]

Definitely following suit on tracking games this year, just for the personal introspection. Love this follow-up to the bigger ep.

[@szbeci]

Great video, thanks a lot for the extra stats. Regarding the Command Zone findings: If someone took a look at their exact data they would find that there were sample groups where fast mana helped people win more games. In commander another huge factor beyond variance is local meta, one playgroup might have totally different playpatterns and reactions than another therefore putting them into one bracket might give a wrong idea. For this reason i loved to hear the different things from your family's meta.

[@TheDracoMaritimus]

The best thing I heard about having fun in Commander comes from the difference between casual and cEDH. Casual is trying to have fun, and hoping to win. cEDH/High Power is trying to win, and hoping to have fun.

[@MrMac1219]

On that last comment; I think you actually learn a lot more about what to change about a deck from the games you win. I'm still trying to figure out my "Ashnod, the Uncaring" deck and just last week had a game I won where I had really been in a board state for 3 turns that could have started doing a lot IF I had haste, but I was relying on finding Thousand Year Elixir for that.

[@Tayahmaryanne]

Last year I tracked games for our four-person weekly pod, and it was interesting. I was tracking for superlatives, not game-altering stats (who was killed first most, who combo-ed out the most, who was our filthy Grixis boi, who won the "owlbear off the top rope" award for highest mana value commander and who won the "ope, Imma sneak right past ya" award for cheapest commander...etc.) This year, I want to track my own games for number of cards drawn and number of lands in play, as well as counting the number of turns I survived, so I can assess how much of my deck I'm seeing and convince myself to run more ramp and card draw. This video was great inspiration and insight, thank you for continuing to be awesome!

[@newdarkcloud]

I think the big lesson here is... that EDHREC hosts enjoy working in Microsoft Excel. Jests aside, this was an interesting watch.

[@devinsdsm]

As one of those 14 Gruul decks in that win column, I can confirm your tracking was unobtrusive and your involvement in the actual gameplay was excellent. You had better grasp and understanding of the board state than almost anyone I've played against, thank you for the awesome game! Also, how do you count turns when I cast an extra turn spell, but then you steal it, copy it, and take two extra turns before I finally win on my extra turn? I challenge anyone to come up with a metric that can track all the craziness that happens in the average commander game.

[@massdiscored7275]

one thing about your numbers is more of personal data and the metas your in, its more of a reflection of you to the game and ppl around you then the game to an overall meta. which still make it interesting to see still cus of small things, popular stuff vs local stuff, shift of favorite decks, and even trying to see if you can speed up games or slow down games with some strategies. that is a heavy emphasis on you. still fun to see that data though.

[@kylekrazymars6726]

Awesome Episode. Enjoyed hearing about the stats. Well executed as always.

[@stone4524]

I just wanna hug you, Joey. You always look so huggable.

[@patonnight]

Purely anecdotal data, but I also have seen the majority of Rakdos and Dimir decks struggle overall to finish games out.

[@Controlqueen31]

And Azorius didn't? I find It is the worst 2 color combination cause there are no good finishers unless you are playing soldier/flying tribal or something like that. The only wincon is usually Aproach of the Second Sun and games are slow.

[@patonnight]

@@Controlqueen31 No. I never encountered those situation nor the card you mentioned. Very different ecosystem I suppose. That's why it was purely anecdotal, everybody personal experience is different.

[@CrabBaskets]

For tracking "fun factor" I wonder if after a game having a notes or comment section where you notate "my deck got to do its thing" or "I felt that the deck performed well and I got to do this cool interaction that I did or didn't know about." And I totally agree that the number of times you play one deck over another does also answer that question, it may be a unwritten intuit, or a vibe that you get and not have to put into a spreadsheet

[@1notdeadfred]

21:11 My solution has been to own a single copy of each dual land I use (none of the og ones, I'm no sadist), and proxy them in my actual decks. This also applies to some of the more expensive cards that I expect to have uses in multiple decks, or that are just expensive enough that I don't feel comfortable taking the actual cards in public, like Volrath Stronghold.

[@hanschristopherson8056]

you and dana’s tracking episode inspired me to track my own games

[@thischannelhasnocontent8629]

Using normal approximation to find a confidence interval for Joey's fast mana statistics -- if Joey's games are a representative sample of commander games (which, to be fair, would be very hard to prove) then we can say with 95% certainty that the winrate of turn-1 fast mana lies between 30% and 62%. It's a wide interval, but it's pretty unlikely to contain 25%.

[@PlagueScholar]

I would actually stay for the lecture about probability, that sounds really interesting! I really wish I had taken more math in college.

[@timothykrzywonski]

I think the fast mana topic indicates why fast mana cab be unfun. Player removal as am obvious and effective strategic solution, but doesn't work without buy in from the table.

[@vasylpark2149]

the answer to how you count a turn is based on who went first and whether it made a full revolution. you don't say if its 8: 45 that is 9. you might say a quarter until 9 but it's still not nine. so, the answer is always it is sometime after 8. You round down. The same thing applies to running track, of if you have ever been in the military during a Fitness test you don't say you ran 6 laps when you only ran 5 and a quarter.

[@1notdeadfred]

18:30 I won't use the data to make sweeping statements, but ffs I'm ALWAYS GONNA BE WEIRD. I have, like, a 2% EDH win rate because weird is always more fun. I always put that first, and while I'd like to win more, I'm just so damn fond of being weird.

[@patonnight]

Nice Gameplay Stats Joey. Now... Let's challenge those stats!!

[@memelorddavid2938]

Man I've been on a wild ride, I started listening to the podcast from the beginning, Matt sure used to be spicy 😂😂

[@JohnGarner-j5e]

i count using a system, give each postion at the table a a letter a,b,c,d and turn followed by turn number start counting turn with the player who goes first. is always "a" ect.

[@DrkshdwG]

How many "I didnt know my deck can do this " did you encounter or other surprise effects?

[@rquer7913]

Hi, thanks for sharing the data... regarding sample size, if only there was an online plattform where people could fill a pre-established form to register their results in a systematic way... not talking about any future feature of EDHREC at all, wink wink... Cheers

[@nigelbarbour1052]

I am loving seeing all this data! One potential factor that you didn't mention was how your own personal decks impact the decks you play against. So the fact you play a lot of black decks can affect the decks your opponents play. For myself, if i see my opponent is playing Dimir, i avoid also playing Dimir, to avoid too much sameness. If this is common, then you playing lots of black decks can contribute to facing more gruul decks, and subsequently you losing to more gruul decks. Maybe youre not losing to gruul decks because they have an advantage, but simply the opponent chose a deck that is different to yours frequently enough to skew the data.

[@towelguy]

I'd like to point out about prices of decks, to consider using the "cheapest print" option for their calculation, a deck fully foiled and secret laired out is for sure more expensive but still the same

[@keanbruhn6210]

I loved both these episodes - i actually totally agree with the playing for the enjoyment - for me its also not important to win just that we have fun - i also found that my decks is not buildt for strong fight - they are buildt for the fun so i do not win much - where i then experience i lose enjoyment is more a saltiness score so to say

[@bingus2464]

On the topic of the decklists changing throughout the year and how that effects the data. Ideally the decklists should remain unchanged to keep the data pure, but as the year goes on, other decks will start playing new cards regardless of if you make any upgrades/sidegrades/downgrades. Assuming power creep, decks with more new cards are more powerful than ones without. One might argue that making reasonable upgrades to decks as the year goes on perserves the integrity of the data more than having consistent deck lists. Bottom line is, if you want perfect data, you're not going to find it anywhere unless you can assure every person playing hasn't added any cards that came out in the past year, good luck with that.

[@jasonritner9662]

I decided to track my games this year and one of the ways I'm keeping track of this type of thing is by marking what I consider "major overhauls" and generally just noting the date of the games I'm playing. If I notice major shifts in data that follow release dates, that may say something. I don't think we need "perfect" data, just data that fits well enough as we should assume people are upgrading decks with each new release.

[@DrukenReaps]

Funny note on the fog thing, Tomer and friends have been bringing up their own use of fog a lot recently xD Granted that is a pretty contained meta among 4-6 folks who rotate a bit and a goofy meta with their self-imposed rules. Seems like even with that though, fogs come in handy.

[@terrymccleland8348]

As a strict budget builder, thank you for mentioning that budget usually has nothing to do with win percentage

[@Crypt4l]

I think about how to tweak my decks after every game, no matter if they win or loose, It's what sparks joy for me.

[@dreyfus37_65]

I’ve heard this phrase a lot and your tangent during the fast mana segment brought it to my mind. “Stats don’t lie, but liars use stats”. Of course someone who doesn’t like playing against fast mana is going to use your data as a weapon it’s statistic to prove their point. And of course they’re going to use the command zone’s data as “an anomaly” or whatever because it doesn’t fit their narrative

[@markrode7245]

Where can i get that mimeoplasm decklist? Made an ooze tribal with him as commander and its fun but pretty subpar. Looking to upgrade him

[@EDHRECast]

Link to all our decklists is in the description of every video! edhrec.com/articles/edhrecast-our-decks/

[@L4TMTG]

Great vid joey!

[@aaronvincent816]

Fun video, very glad that you went through all the hard work with this. I can’t imagine trying to do this AND play actual magic 😂 One question. Do you think your win percentage was affected by you doing such an in-depth study? Like you are focusing so much on different statistics of the game that you miss triggers or little advantages? Again thank you so much for the video and the statistics, this was a very fun analysis of your gameplay and I can’t wait for more!

[@geraldruggles6937]

I've also looked at my decks wins and lost turns. For the most part they seem to last about 2ish turns longer on losses,however I believe I use more interaction than most in my meta. I will NOT go gently into that good night sir!

[@grevedanko]

Awesome episode!

[@adoo765]

See my deck working the way i want its the true fun of EDH to me, even if i lost the game its awesome to have a good turn or a flavor win

[@90kalos1]

A suggestion for tracking turns, just add a note about your position in turn order and the turn position of the winner. That way you can keep tracking turns count on your turn while getting additional data

[@42CZ]

In regards to how the game ends there's a definite difference between me losing a game and someone else winning a game

[@Era5erhead]

SO THIS IS GOSPEL NOW?? No sorry j/k. Really nice video! Thanks for all the effort you put into it and great job pointing out all the caveats. Keep up the good work!

[@Nelphean]

How much mana that is required to make a winning game set may be an interesting way to talk about the impact of fast mana.

[@Dragostorm21]

While true, fast mana does also increase further mana ramp (sol ring on turn 1 into a 4 mana ramp spell further expands the mana lead). The question is super interesting either way (and again, the overall mana needed to win is still a good indicator for the impact of fast mana if you also know how much fast mana boosts the overall mana )

[@rulamagic]

If a lot of people begin tracking games this year, at the end of the year we could consolidate those data and maybe have a good enough sample size to draw more meaningful conclusions.

[@Dragon_Fyre]

I’m curious, as you use the term fast mana what you include in that, as I believe the Command Zone was specifically reviewing stats for just Sol Ring in early turns. I think the real impact from fast mana is when you begin stacking it, having Sol Ring, Mana Crypt and Vault, Moxes etc. where you don’t just have 2 mana up on the other 3 players but rather where you are potentially casting 7-8 CMC spells on turn 2-3.

[@boazverdooner5647]

I think he mentioned tracking sol ring, mana crypt and jeweled lotus appearing on turn 1

[@Dragon_Fyre]

@@boazverdooner5647 I’m surprised at Jewelled Lotus. It’s a one off for your Commander only… I never expect it to be too impactful unless they can Hexproof the Commander. I mostly just like it in my Muldrotha deck since it’s effectively a forever-3 deduction on Commander.

[@ramosroiman]

Hey Joey! Great job explaining this... are you and your team thinking of adding this kind of statistics to archidekt? It would be excellent to incorporate this functionality into your decks and somehow start tracking data on deck performance. Well, maybe it's just me, but I find this kind of statistics fascinating. In many cases, it's subjective, but undoubtedly fascinating.

[@chezpizza3869]

Ok. How many of those decks are running "staple" cards? How much of the deck were made up of "staples"...i.e. what is the uniquiness score of your decks?

[@lithron]

Thumbs up for addressing the fact that Fuji apples and Granny Smith are different things. Both yummy, but not identical.

[@josereinoso6163]

How many decks do you have Joey? Also, did you made an effort to play each of your decks at least once? By doing so, do you think it affect your data collect because you were playing decks you did not want to play and wasting your efforts on decks you do not like playing. Interested to know your thoughts.

[@mrvirtualis]

Thanks for your insights! I did some tracking myself and maybe you like some of my ideas: instead of tracking "fun", I tracked how balanced the game was, subjectively. In other words, was there a deck that was way more powerful or did someone have a non-game? I also tracked whether or not my opponents used tutors (land fetching and others), since I play neither. My take on the what turn did the game end: I used the turn the winning player was in. Most of the times this is the current highest turn, but in case someone wins in someone else's turn, it can be one lower.

[@buggy65x]

Surely EDHREC could make a webpage with drop down selections to normalize the data collected? Then we can all log our games together directly to the database?

[@hoffedemann5370]

You've inspired me to track my own commander games, but how do you deal with ties? Are they wins or losses?

[@Ultinuc]

Why would they be losses?

[@Lyr00]

One of my decks has a surprising 7 game win streak. I think that this number is just luck but some of my friends are terrified when I pull out the deck xD

[@typethe1316]

Hmmm... This makes me think that there should be an app that tracks games data. It would be interesting to see data from a larger data pool.

[@zenalias3131]

ngl, I chucked at the idea of calculating p-value.

[@gnomersy1087]

I do think that you can have a fun game without winning but I generally don't find games where I never had a chance to win very fun. Not everyone agrees with me on that though. I also think looking at your deck after losses and wins is a valuable exercise but I don't think you necessarily do that every game. I do think that if you find a glaring hole in your deck after a game that is something you ought to fix. For example, I got ran over by a go wide deck and realized my deck had zero wraths in it so I had no draw that would save me. Now I could just as easily have had the right number of wraths and due to poor draws on my part or good draws on my opponents still have gotten steamrolled but not have needed a change. I find your data tracking exercise interesting obviously if you were a scientist doing a study there'd be different expectations but even as a backyard science approach this is really fun info to get to see as long as we all take it with a few grain of salt.

[@Crow_Saw]

Joey: Don't take these numbers as gospel. Some guy on the Internet: Got it! Fast mana wins 100% of the time! I'm going to play 30 lands and 70 Sol Rings!

[@henrivalencia21]

haven't seen every video you posted. But what is the story of the scroll of fate in your background?

[@EDHRECast]

One of our first previews! :)

[@ElDocBruh]

My sample size is "I wish I could play this much MTG" 😮‍💨

[@SamuraiJM]

I see Mimeoplasm, I click video.

[@johnnyborbon5085]

I get the first part about recording average won game length and average lost game length but tbf i doesn't tell me anything unless you can prove statistical differentes between the 2 game lengths, comparing averages in this specific case doesn't do much for me

[@Durtaz]

Ngl just slightly disappointed that this wasn’t including a college lecture on probability mathematics. Great video though!

[@heavyyield]

Can you do stats on turn order? Like going first vs going last?

[@joedoe7572]

He had stats about going first in the other video. I really want to see stats about going last too though

[@VincentWolfeye]

I think the "fun" metric is a good one to have. Maybe just a scale from 1-5. 1 - I didn't like the game 3 - It was the kind of game I hoped for. 5 - the game exceeded my expectations. Maybe that metric should come with the caveat that the result didn't take into account what promted the game to be good or bad. A game in which ones deck got manaflood or screwed could be as valid of a bad game as a not meeting the expectations one had. Maybe a better metric could even be 1 - not meeting expectations 3 - meeting expectations 5 - preforming over expectations Just ideas...

[@narvuntien]

As we have definately learnt recently people are very very bad at understanding probability and interpreting it. Otherwise gambling wouldn't exist. Your results will get a lot better with 5 years of data, I think your best option is to do it for 5 years. The probability stuff I do get a bit angry about is people playing fetchlands in mono colour or off colour fetches in decks that are not landfall, sac or have some specific synergy with them, just because it "thins the deck". Probably the only time my deck building brain got tilted seeing a Mono white deck with evolving wilds.

[@RyanTraher]

Wait… what if *I* want to see you do p value calculations in front of me? I went to grad school for econometrics, it has to be good for something.

[@taylorarnold880]

I've tracked all the commander games I've played for the last 2 years. That would be 0. 0 games. With 23 decks. Because I have no playgroup or LGS.

[@1notdeadfred]

14:15 Fuji Apples are superior. Fight me lmao

[@Deluxe497]

Granny Smith guild!!

[@grampaw5605]

Applying statistics to a chaotic system, like MTG, has it's drawbacks doesn't it?

[@ajaydassani9931]

I'm curious did you switch cards in your decks in the past year?

[@EDHRECast]

25:01

[@brendans1983]

If tracking all my games for the year has taught me one thing, it is that all that data means nothing without all the other variables 😂

[@RazgrizAce67]

I'm thinking about trying to start a Discord channel for all those folks like us who log our games to start putting our data together. So we can start understanding bigger picture stats. Would you be interested in that?

[@ssriceboat]

More data!? *drools* IN A GOOD WAY LIKE HUNGRY FOR DATA

[@drunkgerbil9962]

Bro Red... is in pain..

[@HalfVastGamer]

Don't you mean: The Bonus Dada

[@gabrielharvey956]

I got a very different view of fast mana than you do , sol ring is good , but mana crypt I'd say 85 % of thr time som1 turn 1 mana crypt they win if it stays like 2 3 turn 2 mana for 0 is ban worthy , it's like beginning with 2 turn in advance its totaly unfun and pay to win

[@brennanclement8582]

GIVE US THE P VALUE CALCULATIONS!

[@michaelcalhoun5585]

More probability metrics!!!!

[@turfinat0r]

No no no dont let them know about Grumgully

[@BubbaGump1211]

Give me those P values!

[@BubbaGump1211]

For turn count, could you start at 1 and count by 0.25 for each players turn? Example; I go second and its turn 3, so the turn count is 3.25

[@deadtv2712]

Ok, although you keep dismissing your stats as “these is just my experience” it still holds some value if your statements are all true. Yes the sample size may be small for some it still represents some valuable information that holds SOME value

[@eridevonta1384]

Did Magic arts made you call your ex and made them came back to Magic? Hilarous lol

[@cedarbobedar7223]

fwiw if you *wanted* to get more accurate and/or detailed information, it seems like recording each game with a top-down camera and data-mining it after the fact, potentially with some support so it's not a monumental task, would make sense. It sounds like an "intern" kind of task at the scale of a magic player but I imagine there are probably some people out here on the internet that enjoy and are quite good at doing that kind of post-game analysis, and frankly it's not so different than the types of post-game analysis that are applied to professional broadcast sports. Actually, doing that and editing it out to follow specific commanders/decks would make for interesting youtube videos. Ideally you would have a group of people playing decks that overlap against people who aren't from that group at a relatively consistent power level so it wouldn't be a small endeavor, but it could be really cool.

[@cedarbobedar7223]

like present the data in context, in 20ish minute videos featuring each deck in a set of X decks that are played in different variants by maybe two or three people over a few months of filming time, with face-to-camera bits in between segments talking about changes they made to the deck and why and then just summarize each game and offer the stats as well as updated averages as you progress through the episode - okay please do it - now I want to watch it lol