People’s Average Game Speed

[materpillar]

We had some discussion of game speed pop up over in the random card of the day thread. I actually record most of my games because I'm a weird statistics nut and I'm curious to know how they match up to other people's game length expectations.

73 games recorded
Games ending turn 5-6 : 8/73 (11%)
Games ending turn 7-10 : 35/73 (48%)
Games ending turn 11-15 : 24/73 (33%)
Games ending turn 16-18 : 6/73 (8%)
Average length 10.5 turns

I don't expect that anyone else has been recording in quite so specific detail but I'd be curios to know roughly what turn people expect games to be finished by in their meta. I say "finished" because I'm referring to when the game is basically over not explicitly over. When does the stasis/control player lock-up the game, not when do they finish beating people to death with a Mutavault or what have you.

[pokken]

Most of my games are over by turn 8 or so these days. Someone is doing something must-answer by turn 4-5 every game.

[PrimevalCommander]

My group is a bit weird, since it is rife with 1+2 combos hidden inside generally weak decks. So either they get lucky and combo out turn 6, or they don't and I coast to victory by turn 8. But honestly I never pay attention to actual turn count. I probably should, but I don't get enough games in to get any reliable data. Mainly by how long the game was. Average game length is about 1-1.5hrs right now I would say. Which seems to be pretty good for how my decks are built right now. I have a control deck that wants to push the game to turn 10+ but if there is little interaction with me, I'll combo out before then. I like to think I live in the 75% optimized zone, but some decks are closer to 85%.

[ISBPathfinder]

For me and the people I tend to play against it comes down to a question of actions in a turn. As you get more actions in a turn turns usually go longer. This often comes from triggers, cheap cantrips, draw, and ramp. Often times turn one / two are the fastest but that isn't to say that later turns are always longer.

[Treamayne]

I'd be curios to know roughly what turn people expect games to be finished by in their meta. I say "finished" because I'm referring to when the game is basically over not explicitly over.

While I don't have a record, I can tell you my last five "casual" MTGO games were roughly (based on how many turns I had - which may be +/-1 from the turn number of the winner):
Functionally over: 5 Explicitely over: 8
Functionally over: 3 Explicitely over: 10 (Mogis anti-creature stax w/ LD "Casual" - durdled until he could find Eldrazi)
Functionally over: 6 Explicitely over: 6
Functionally over: 11 Explicitely over: 13 (Best game in over a year)
Functionally over: 5 Explicitely over: 7

[DirkGently]

I'd be curious to actually track this. Currently I think we tend to have a pretty wide range. Jon is typically trying to do something pretty powerful that either gets answered and then he sulks and the game goes on forever, or he run away with the game and wins pretty quickly. His infect deck I think will commonly win on turn 6 or so. But we've also had plenty of games that must have gone to turn 20 or more.

[TheAmericanSpirit]

I think it really depends on what the table composition contains. You get 2+ grindy control decks in a pod and things can get glacial real quick and stay that way for a good while. Whereas if you throw together Torbran, Thane of Red Fell, Yusri, Fortune's Flame, Raffine, Scheming Seer, and Liesa, Shroud of Dusk into a game and we're almost spending more time on shuffling and mulligans than playing.

Certain cards can also extend/shorten the lifepspan of a game (for better or worse). For example, every single time I see Humility in a game, I can reliably count on at least 4 more turn cycles than the median EV. Sulfuric Vortex going unanswered will melt 40 life down in a jiff. Whether those cards stick and if so for how long is a big factor in my experience.

Obviously meta composition is also a factor. Turn 3 is Go Time for cEDH whereas the right pile of precons could theoretically sling cardboard for hours. Are there house rules or other accoutremont, like archenemy or planechase? That's another variable that can add or subtract depending on a myriad of circumstances.

Basically what I'm saying is that Turn Count itself is not enough. We have to isolate and identify the variables, chart the data, and then look for statistically signifcant patterns. Only with a thorough grasp on the HOW will we ever truly understand the underpinnings of WHY a game takes as long as it does.

As for "finished", I think that's less useful than measuring the actual game length because it does not acknowledge the possibility of a comeback. A timely wipe, an impromptu combo, a clutch Spore Cloud, etc can really change the course of a game. It's not over until it's over.

[materpillar]

For me and the people I tend to play against it comes down to a question of actions in a turn. As you get more actions in a turn turns usually go longer. This often comes from triggers, cheap cantrips, draw, and ramp. Often times turn one / two are the fastest but that isn't to say that later turns are always longer.

I was asking less about average minutes spent per turn and more "on what turn does the game seem to end in your metagame". I'm curious to get a vague feeling on how my metagame's speed is compared to other users here. For example, it appears @pokken has a meta that's probably 2-3 turns faster than mine, which is a ton.

Basically what I'm saying is that Turn Count itself is not enough. We have to isolate and identify the variables, chart the data, and then look for statistically signifcant patterns. Only with a thorough grasp on the HOW will we ever truly understand the underpinnings of WHY a game takes as long as it does.

I'd guess that average game length very strongly correlates with a metagames powerlevel. If someone says "our games are usually over on turn 6", I'm going to be busting out a very different deck than if someone says "our games are usually over turn 12". I'd guess knowing your average + standard deviation would give a way better picture. That being said. I can give you an extremely accurate metagame breakdown of my last 2 years of play including things like each players deck choices, winrates per deck / against deck and average game length per player per deck. I'm very aware that I'm a bit of an outlier in terms of enjoying data and its recording. As such I'm guessing the best comparable data I can get here is "my median game length seems to be roughly X". If you've got an in-depth statistical analysis of your metagame I'd be hyyyyyyyyyped to look at it.

@3drinks You play most frequently on MODO right? Does it record the games? Can you look at what some of your recent game lengths were?

As for "finished", I think that's less useful than measuring the actual game length because it does not acknowledge the possibility of a comeback. A timely wipe, an impromptu combo, a clutch Spore Cloud, etc can really change the course of a game. It's not over until it's over.

I'd simply argue that those games weren't finished then. My "finished" term was mostly a quick and dirty way to avoid things like @DirkGently's Phelddagrif's deck massively inflating estimated end turn length. That deck basically wins and then takes another 6+ turns to beat the last poor soul to death with a hippo.

I'd be curious to actually track this. Currently I think we tend to have a pretty wide range. Jon is typically trying to do something pretty powerful that either gets answered and then he sulks and the game goes on forever, or he run away with the game and wins pretty quickly. His infect deck I think will commonly win on turn 6 or so. But we've also had plenty of games that must have gone to turn 20 or more.

This actually sounds pretty similar to my meta except my "Jon" is extremely aware he's doing degenerate things and doesn't mind getting mauled into oblivion.

[3drinks]

@3drinks You play most frequently on MODO right? Does it record the games? Can you look at what some of your recent game lengths were?

Doesn't look like I have anything in my past games most likely the data was wiped when Daybreak Games took over MODO. Even when I did have some though, I don't think these were an adequate sample because of the sheer dearth of differing power levels - there's always inevitably the "casual" player that insists on being a hipster that refuses to play "so-called best" cards that would have more than a 4% utilization on edhrec - you know the type, they always 1-3 their FNM and complain about people playing actually good cards - and inevitably mixed in with The Sweaty Tryhard™ playing their all foil Edric deck like they think they are clever. While also I've had 1v1 games with a friend where we gotta be hitting t8-10 because I'm hardcasting Razaketh or Villis. What I'm saying is, this isn't a metric that adds anything to the discussion I think.

In paper, there's always this player with some warhammer card, I can't remember it's name but it's a 3/5 in esper that makes free bodies or something, and combined with the orzhov sherrif that makes free treasures and I think draw? and nobody does anything to pressure those resources but me, but I can't get through it's wall of permission. The other two players don't interact, or even meaningfully contribute to the board so even my debatably high interaction suite doesn't matter against free resouces + countermagic. What I'm saying, is we don't have a balancing point yet and as long as decks aren't balanced, measuring turn rates don't matter. The guy with the 3/5 then finishes on lab man.

[Dragonlover]

I'd have to actually track it, but my gut is 7-10 for the majority of our 4 player games, which given I've always thought if materpillar showed up at our shop he'd slot right in feels about right. I know he doesn't because I've never had my board turned into Goblins then stabbed.

Dragonlover

[Avacyn Believer]

I would love to give you an accurate answer but I never thought to track turns either. If I had to guess I'd say same others, 8 to 10 turns per game? I am curious now, but not sure I'll remember to keep records

[PrimevalCommander]

I played 10 games last weekend in about 12-13 total hours. Average game time was about 1hr 20min across that time period. And every game we played had a good cadence with minimal board stalls. Plenty of board wipes, but not too many stalls for more than a turn or two. One of the most productive weekends of magic I have had. Though we only had about 4 games with 4 players. So the majority were 3 player games. Not ideal, but that's what we could get.

Quickest win I think was turn 7, in 3 player game where one player got mana screwed and I took over quite quickly. Longest win I didn't track but was well past turn 10. 4 player game with good back and forth. Good variance, but by turn 8-10 the game is heating up... or cooling down if the control player finds a loop on a board wipe and locks down the rest of the table. Either way, we are positioning for a win by turn 9-10. Whether it happens depends on how much removal was drawn and now many decks with blue are at the table. Actually we didn't see all that much blue all day, which probably contributes to the modest game lengths.

[Jemolk]

In my meta, games regularly go long. Like, really long. As in, 3 hour games aren't uncommon, let alone unheard of, long. Now, this is mainly because the games go every which way, with a wide variety of interaction and everyone being in the lead at some point, which I absolutely consider to approach the Platonic ideal of an EDH game, but there's no denying that they take a while. Turns-wise, too. I've never actually counted turns, but it's relatively rare that I don't hit 8+ natural land drops in a game, and most of those where I don't involve mana screw or getting taken out early. 15-20 natural land drops is hardly unheard of, though hardly a constant -- I'd attribute some of that to missing land drops, though. Games where I have my Thawing Glaciers are games where, if I'm in the game all the way to the end, I generally end with 18+ lands on the field with absolutely minimal ramp, if any. I have a few decks that are for significantly faster games, like mono-Red Chandra Tribal, but even then, turn 8-10 is close to the expected end point. The turns end up being much faster, too, though.

[Avacyn Believer]

I've been keeping turn count with my main playgroup yesterday and our average was 7 turns. Some games were faster, but not many were longer. There are only three of us playing though, so that might affect it. I'll try to remember to keep a record when I play in a regular game of four.

[aliciaofthevast]

Well, I know my games with Aron have been going long. But I'm the control deck so that's how it's supposed to be right? I find I've exhausted your firepower and have started taking over around t12-15. Usually on the back of a Sun Titan or Reveillark. I don't actually try to win until I'm sure I've exhausted your flurry though because, well, I'm the control deck, I've got time....

This is on Magic Online, so I've been clocking around 50-65min/game I think if I have to guesstimate! We're not at full power decks (maybe at a 7? 😜) and it's 1v1 vs a tuned Zurgo Bellstriker. I keep the power level roughly even so I don't overshadow their deck 😎

Maybe I'll make a YouTube channel and upload game replays. That sounds like too much work for no benefit though 😳

[Moxnix]

On mtgo a decade ago it was turn 8-12 average now a days its more like turn 5-8 I don't know how reflective that is of the paper meta but lots of expensive staples are cheaper online and now a days even at non CEDH tables I expect everyone to be making plays turn 1-2