Best Win: T1 Dryad Arbor, T5 Rasputin Dreamweaver, T6 Iona...
Worst Loss: Opponent got T4 Nightscape Master....
Best Game: Both Opponent & I were at 1 life. I had Vigor and other creatures in play (including Escape Artist), Opponent had Blazing Archon in play....Opponent finally got Bringer of Red Dawn to end game, Took Vigor and attacked for the win....
Yes WhiteTig3r was playing the classic mono-blue fish deck it was nearly an identical build to the one LE played a few weeks ago, the only difference I noticed was the inclusion of Mana Drain.
Uhm, I noticed a strange thing: I NEVER use any other avatar than Higure, the Still Wind. Why am I looking like a Serra Angel in those games?! Now I wonder, do I look like Serra Angel to everyone? I'm a sneaky ninja not a sexy angel!!
Topic related: My deck was casting X=20+ Banefires constantly but you ruined my gameplan... not with Lightning Rift though. It was your Withered Wretches that destroyed me. You shut down my Crypt and thus shut me down. They were good games overall.
Tuktuk Grunts is not better than Burst Lightning at a statistically significant level. As I have said in past articles (and I think in this one as well), that requires a different test altogether, and there are so many combinations of cards that it isn't worth doing on the whole.
And it truly is amazing how bad players are with removal spells.
All the time I had that silly feeling: "Am I too old to play Magic?". I think you would understand me very well. I had to overcome that feeling and that day at the shop I saw that I was not too old to play Magic at all.
true...though i can also think of many creatures i would rather have than the grunts...unless im allies all the way, i dont think i would even push him as a 23rd..
Removal, like Burst Lightning, is also one of the most misplayed type of cards. Many inexperienced and new players use removal right away and often on insignificant targets. Creatures on the other hand are easier to play. Knowing when to attack or block with a creture is significantly easier than knowing when and what to target with removal.
If large numbers of players misplay Burst Lightning, does that make Burst Lightning a bad card? No...
If large numbers of players misplay Burst Lightning, does that make its percentage in this study go down? I'm guessing yes...
"I stayed at the shop for more than four hours and had great time. All the other players accepted me into their group (they didn't had to) and they were all very nice to me. I promised them to show up for the coming Worldwake prerelease and release events and left the shop with many questions in mind:
Why-o-why didn't I do this before? Why-o-why didn't I move my fat a** and didn't go to a prerelease before? And why-o-why did I allow myself to forget the pleasure of real Magic with real cards?"
Welcome back. Enjoy the cardboard part of cardboard crack!
Your daily breakfast? I bet you would be a little more fit if you skipped the bacon. Is there not some biased news channel you could absorb information from instead?
I just cant see how a card that is really only mediocre/alright in one match up is the best red card over much more important and splashable cards that red offers. "Statistically" it makes no sense. Because there has never been a single moment of my life when I drew a burst lightning and thought.. "Man, I really wish this was a Tuk-Tuk Grunt."
oh i had a bant spread 'em homebrew, though i was referring to the U/G land control deck the one guy called a netdeck... spread em is great to play...not much to play against
also note he have 1 bloodghast in the grave,
so any land would give him the win , or gatekeeper, or hexmage , any feast , nighthawk,
on the other hand if he pulls another blade , or bloodwitch , or mind sladge you have a shot
Small problem with that - since everyone and their mother either uses the price list that Heath uses or steals it from him, most bots out there are buying Foil Chalices for less then the non-foil versions. For no apparent reason, a foil Chalice is worth essentially LESS than a regular one.
But William makes a note to say that cards that appear less than 20 times are not counted.
However if a card that is played 140 times is 'weighed' as the same as a card that is only played 30 times, it's possible some of McFish's concerns are relevant.
The problem with counting the cards present in a deck as McFish suggests is that MTGO replays don't show you what is in their deck, just what was played.
Good comments but you still haven't said what the point of this series is! I still don't know how good tuktuk grunts actually is because there are too many confounding factors to reach any conclusions based on the data.
1) I'm not just relying on my experiences here, I'm relying on general discussion around the cards, which is admittedly based on anecdotal evidence but the experiences of hundreds of players, many of whom are very good. The consensus of this discussion is that tuktuk grunts is not a very good card, and certainly not a game winner over geopede or lightning. I've played with the cards enough to conclude that the consensus is not too far from reality. I'm not going to argue that my results are 100% accurate, I'm saying that you have no basis to claim that your results resemble reality or the 'bigger picture'.
2) Well yeah, ondu cleric is a very important card in an allies deck. How does "55% of games where player played ondu cleric resulted in win" tell you that though? It doesn't, because 55% is just a number and not evidence. It doesn't account for the archetype or matchup at all.
3) You don't need a larger sample size. That's not your problem. Your problem is you've yet to make a conclusion, or even a meaningful analysis. Are you trying to say tuktuk grunts is a good card, a must-include or a high pick? I don't think you've shown that at all, like I said, by your analysis iona would be the best card ever printed. That doesn't even conform to common sense. The fact that your results are not even adjusted to something as simple as mana cost shows that they are no better than mere opinion. You're merely extrapolating from meaningless numbers.
In summary, various posters have pointed out the critical flaws of your methodology, yet you insist that your results have merit. I'm finding it very difficult to draw anything from these numbers to improve my play. Perhaps you can build a pick order out of these statistics and tell me how that goes.
It certainly does not correct things, and "cards drawn" or "cards played in deck" would be a lot better--though the latter would need a huge n to pick up on things.
1) Have you considered that the sample of games you have played is very small, and thus YOUR results might not be accurate? That's one of the take home points of these articles. While I don't want to discount your games as being irrelevant (they are not), you need a bigger picture, and you simply can't get that out of the number of games one individual plays.
2) Ondu Cleric did not come back as statistically significant, so its result does not deter me in the least. But it does go to show you that if you have a lot of allies in your deck, it's not the end of the world to run it.
3) A correction (CAPS are mine): "Informal analysis allows us to say 'ok I THINK card A is better than card B because of C D E.'" There's a big difference there. Statistics won't tell us this kind of stuff without an even larger study, but I can say with a lot of certainty when a card is good or not. And, as we have seen, not all of these cards are obvious.
What can we actually learn from this analysis? You have gone 4 articles without actually answering this, without drawing any kind of valid conclusion backed up by the facts.
Informal analysis allows us to say "ok card A is better than card B because of C D E". You yourself said statistical analysis doesn't tell us which card is best. So what does it tell us? It tells us what card might be better than we'd expect? We guess at what the numbers actually mean? How is that better than informal analysis? I'm a mathematics major, so I like to let the numbers talk, but without context, without understanding, mere numbers are useless.
What errors have we actually found? The 'assumptions laid out' have completely detached the analysis from the game itself. "If all cards have the same chance of getting played", "if all cards belong in all decks", "if sideboard cards get brought in every matchup", etc. These are invalid assumptions with huge sampling bias that make the results utterly meaningless.
So what does tuktuk grunt's 63% actually tell us about the card, how it should be included and played? If it's not the best card, or even a good card compared to other red commons, why do we even bother giving it a number?
Don't look at makindi shieldmate (even though 50% is way too high for shieldmate), look at ondu cleric. You yourself noted that the statistic is misleading. The moment you discovered that result you should've stopped and reevaluated your entire project.
You've only addressed it by equating "mana screw" with "mana flood".
Take the text output of the games, pipe through awk, and then do a histogram of losses by number of lands on the table.
As an experiment, I tried to figure out the card that would do the best in your rankings to illustrate the flaws we've been discussing.
Realm Razor:
* Requires you to have 6 lands
* Requires you to have 3 colors of mana available
* Is rarely played when you're losing
I think these articles are incredibly interesting and there's a lot to be said for thinking about and discussing the methodology. I appreciate not only the time you've put into this, but also the way you've discussed people's comments. Very good stuff.
That week was terrible for me. I came with a really subpar red/green elves deck and it lost very quickly to MagicCardMaker.
i never saw the mana drains, but then i also never saw force in LE's version when I played against him many weeks ago. But thats for clarifying
Best Win: T1 Dryad Arbor, T5 Rasputin Dreamweaver, T6 Iona...
Worst Loss: Opponent got T4 Nightscape Master....
Best Game: Both Opponent & I were at 1 life. I had Vigor and other creatures in play (including Escape Artist), Opponent had Blazing Archon in play....Opponent finally got Bringer of Red Dawn to end game, Took Vigor and attacked for the win....
Thanks for the primer...
Yes WhiteTig3r was playing the classic mono-blue fish deck it was nearly an identical build to the one LE played a few weeks ago, the only difference I noticed was the inclusion of Mana Drain.
You were a sneaky ninja last night when I played you.
Uhm, I noticed a strange thing: I NEVER use any other avatar than Higure, the Still Wind. Why am I looking like a Serra Angel in those games?! Now I wonder, do I look like Serra Angel to everyone? I'm a sneaky ninja not a sexy angel!!
Topic related: My deck was casting X=20+ Banefires constantly but you ruined my gameplan... not with Lightning Rift though. It was your Withered Wretches that destroyed me. You shut down my Crypt and thus shut me down. They were good games overall.
LE
Tuktuk Grunts is not better than Burst Lightning at a statistically significant level. As I have said in past articles (and I think in this one as well), that requires a different test altogether, and there are so many combinations of cards that it isn't worth doing on the whole.
And it truly is amazing how bad players are with removal spells.
Thank you Pete.
All the time I had that silly feeling: "Am I too old to play Magic?". I think you would understand me very well. I had to overcome that feeling and that day at the shop I saw that I was not too old to play Magic at all.
LE
Do you have anything of value to add regarding the statistical discussion?
Because, flawed as it is, at least it's unique. Your post, not so much.
true...though i can also think of many creatures i would rather have than the grunts...unless im allies all the way, i dont think i would even push him as a 23rd..
Removal, like Burst Lightning, is also one of the most misplayed type of cards. Many inexperienced and new players use removal right away and often on insignificant targets. Creatures on the other hand are easier to play. Knowing when to attack or block with a creture is significantly easier than knowing when and what to target with removal.
If large numbers of players misplay Burst Lightning, does that make Burst Lightning a bad card? No...
If large numbers of players misplay Burst Lightning, does that make its percentage in this study go down? I'm guessing yes...
Great article.
And as for:
"I stayed at the shop for more than four hours and had great time. All the other players accepted me into their group (they didn't had to) and they were all very nice to me. I promised them to show up for the coming Worldwake prerelease and release events and left the shop with many questions in mind:
Why-o-why didn't I do this before? Why-o-why didn't I move my fat a** and didn't go to a prerelease before? And why-o-why did I allow myself to forget the pleasure of real Magic with real cards?"
Welcome back. Enjoy the cardboard part of cardboard crack!
Your daily breakfast? I bet you would be a little more fit if you skipped the bacon. Is there not some biased news channel you could absorb information from instead?
I just cant see how a card that is really only mediocre/alright in one match up is the best red card over much more important and splashable cards that red offers. "Statistically" it makes no sense. Because there has never been a single moment of my life when I drew a burst lightning and thought.. "Man, I really wish this was a Tuk-Tuk Grunt."
oh i had a bant spread 'em homebrew, though i was referring to the U/G land control deck the one guy called a netdeck... spread em is great to play...not much to play against
also note he have 1 bloodghast in the grave,
so any land would give him the win , or gatekeeper, or hexmage , any feast , nighthawk,
on the other hand if he pulls another blade , or bloodwitch , or mind sladge you have a shot
No guff
Small problem with that - since everyone and their mother either uses the price list that Heath uses or steals it from him, most bots out there are buying Foil Chalices for less then the non-foil versions. For no apparent reason, a foil Chalice is worth essentially LESS than a regular one.
McFish makes a good point.
But William makes a note to say that cards that appear less than 20 times are not counted.
However if a card that is played 140 times is 'weighed' as the same as a card that is only played 30 times, it's possible some of McFish's concerns are relevant.
The problem with counting the cards present in a deck as McFish suggests is that MTGO replays don't show you what is in their deck, just what was played.
Good comments but you still haven't said what the point of this series is! I still don't know how good tuktuk grunts actually is because there are too many confounding factors to reach any conclusions based on the data.
1) I'm not just relying on my experiences here, I'm relying on general discussion around the cards, which is admittedly based on anecdotal evidence but the experiences of hundreds of players, many of whom are very good. The consensus of this discussion is that tuktuk grunts is not a very good card, and certainly not a game winner over geopede or lightning. I've played with the cards enough to conclude that the consensus is not too far from reality. I'm not going to argue that my results are 100% accurate, I'm saying that you have no basis to claim that your results resemble reality or the 'bigger picture'.
2) Well yeah, ondu cleric is a very important card in an allies deck. How does "55% of games where player played ondu cleric resulted in win" tell you that though? It doesn't, because 55% is just a number and not evidence. It doesn't account for the archetype or matchup at all.
3) You don't need a larger sample size. That's not your problem. Your problem is you've yet to make a conclusion, or even a meaningful analysis. Are you trying to say tuktuk grunts is a good card, a must-include or a high pick? I don't think you've shown that at all, like I said, by your analysis iona would be the best card ever printed. That doesn't even conform to common sense. The fact that your results are not even adjusted to something as simple as mana cost shows that they are no better than mere opinion. You're merely extrapolating from meaningless numbers.
In summary, various posters have pointed out the critical flaws of your methodology, yet you insist that your results have merit. I'm finding it very difficult to draw anything from these numbers to improve my play. Perhaps you can build a pick order out of these statistics and tell me how that goes.
Thanks again for the comments.
@Paul: It seems that I did misunderstood your first comment.
@anonymous: You're right indeed and Yusuf really is an amazing guy.
@Shard: Playing Spread'Em is a lot of fun. Just be careful where you play the deck; playing it in the casual room won't make you any new friends :)).
Thanks again for the comments.
LE
lol man, you are turkish. i lived in turkey for a long time and played magic there. the guy in the picture is yusuf. he breathes magic.
It certainly does not correct things, and "cards drawn" or "cards played in deck" would be a lot better--though the latter would need a huge n to pick up on things.
Some comments:
1) Have you considered that the sample of games you have played is very small, and thus YOUR results might not be accurate? That's one of the take home points of these articles. While I don't want to discount your games as being irrelevant (they are not), you need a bigger picture, and you simply can't get that out of the number of games one individual plays.
2) Ondu Cleric did not come back as statistically significant, so its result does not deter me in the least. But it does go to show you that if you have a lot of allies in your deck, it's not the end of the world to run it.
3) A correction (CAPS are mine): "Informal analysis allows us to say 'ok I THINK card A is better than card B because of C D E.'" There's a big difference there. Statistics won't tell us this kind of stuff without an even larger study, but I can say with a lot of certainty when a card is good or not. And, as we have seen, not all of these cards are obvious.
You're dodging the question.
What can we actually learn from this analysis? You have gone 4 articles without actually answering this, without drawing any kind of valid conclusion backed up by the facts.
Informal analysis allows us to say "ok card A is better than card B because of C D E". You yourself said statistical analysis doesn't tell us which card is best. So what does it tell us? It tells us what card might be better than we'd expect? We guess at what the numbers actually mean? How is that better than informal analysis? I'm a mathematics major, so I like to let the numbers talk, but without context, without understanding, mere numbers are useless.
What errors have we actually found? The 'assumptions laid out' have completely detached the analysis from the game itself. "If all cards have the same chance of getting played", "if all cards belong in all decks", "if sideboard cards get brought in every matchup", etc. These are invalid assumptions with huge sampling bias that make the results utterly meaningless.
So what does tuktuk grunt's 63% actually tell us about the card, how it should be included and played? If it's not the best card, or even a good card compared to other red commons, why do we even bother giving it a number?
Don't look at makindi shieldmate (even though 50% is way too high for shieldmate), look at ondu cleric. You yourself noted that the statistic is misleading. The moment you discovered that result you should've stopped and reevaluated your entire project.
You've only addressed it by equating "mana screw" with "mana flood".
Take the text output of the games, pipe through awk, and then do a histogram of losses by number of lands on the table.
As an experiment, I tried to figure out the card that would do the best in your rankings to illustrate the flaws we've been discussing.
Realm Razor:
* Requires you to have 6 lands
* Requires you to have 3 colors of mana available
* Is rarely played when you're losing
I think these articles are incredibly interesting and there's a lot to be said for thinking about and discussing the methodology. I appreciate not only the time you've put into this, but also the way you've discussed people's comments. Very good stuff.