Blitz Team Stats

[Fleck]

For those who are curious (or just really bored, like me), this is a composite of our overall team records and stats after four leagues. Tournament games are not factored into these statistics. I'll probably add more detailed stats and whatnot later.

UPDATED: 5TH LEAGUE

Overall team Win/Loss ratios:

#1: Loveless: (26-9-10)
#2: Grizzly Bears: (25-13-7)
#3: Nightshades: (22-15-8)
#4: Mad Experiments: (22-18-5)
#5: Mercenaries: (18-17-10)
#6: Magic Missiles: (16-22-7)
#7: Sublimely Magnificent 7: (14-24-7)
#8: Guests Who: (12-25-8)
#9: Marlboro Mayhem: (4-3-2)

Overall Points For:

#1: Grizzly Bears: 255
#2 (tie): Sublimely Magnificent 7: 201
#2 (tie): Guests Who: 201
#4: Mercenaries: 195
#5: Loveless: 180
#6: Magic Missiles: 171
#7: Nightshades: 160
#8: Mad Experiments: 159
#9: Marlboro Mayhem: 54

Overall Points Against:

#1: Marlboro Mayhem: 50
#2: Loveless: 114
#3: Nightshades: 136
#4: Mercenaries: 164
#5: Mad Experiments: 171
#6: Grizzly Bears: 190
#7: Magic Missiles: 205
#8: Guests Who: 237
#9: Sublimely Magnificent 7: 240



I might be off by a couple of digits here and there, but I think overall, these standings are accurate. Let me know if you guys want to see more detailed team stats and comparisons.

[wyvernxk7]

I love this, obviously

[Fleck]

Since the last list didn't contain any tournament data, here's all the tournament stats up to the third tourney. I'll update this list after the conclusion of the 4th tourney.


UPDATED: 4TH TOURNAMENT

Most Tournament Games Played:

#1: Mercenaries (14 games)
#2: Grizzly Bears (13 games)
#3: Loveless (11 games)
#4 (tie): Nightshades (10 games)
#4 (tie): Mad Experiments (10 games)
#5 (tie): Guests Who (6 games)
#5 (tie): Magic Missiles (6 games)
#6: Sublimely Magnificent 7 (5 games)

Overall Tournament Record:


#1: Mercenaries (9-5)
#2 (tie): Grizzly Bears (8-5)
#2 (tie): Loveless (8-3)
#3: Nightshades (6-4)
#4: Mad Experiments (4-6)
#5: Guests Who (2-4)
#6: Magic Missiles (1-5)
#7: Sublimely Magnificent 7 (0-5)


Average Points For:

#1: Mercenaries (5.1)
#2: Grizzly Bears (4.7)
#3 (tie): Mad Experiments (4.6)
#3 (tie): Loveless (4.6)
#4: SM7 (4.4)
#5: Guests Who (4.2)
#6: Magic Missiles (3.8)
#7: Nightshades (3.6)


Average Points Against:

#1: Loveless (3.1)
#2: Mercenaries (3.6)
#3: Grizzly Bears (4)
#4: Mad Experiments (4.3)
#5: Nightshades (4.8)
#6: Guests Who (5.8)
#7: Magic Missiles (6.2)
#8: SM7 (6.8)

Highest Average Finish:

#1: Mercenaries (2.5)
#2: Loveless (2.8)
#3: Grizzly Bears (3)
#4: Nightshades (3.5)
#5: Mad Experiments (4)
#6: Guests Who (4.3)
#7: Magic Missiles (5)
#8: SM7 (5.25)



Notes:

Some of this data may seem a bit skewed, but that's because we've had a couple of wonky tourney games. The Guests Who did not qualify for the first tournament, so that's why they're the only team with an even number of played games. In the third tourney, Mad Experiments demolished Nightshades 14-1, which results in the weird Points For and Points Against stats for those two. I decided to use Average Points For/Against rather than total because we have an unequal number of games played, so I felt that averaging the results produced fairer and more accurate stats. For Highest Average Finish I averaged 1st through 4th places, and then assumed that all teams that were eliminated in the first round finished 5th.

Also, Mercs suck.


4th tournament notes:

Well, the double-elimination format certainly helped clear up a lot of our ties, so that now everyone's pretty clearly spaced out. Unfortunately, this has had the effect of putting the Mercs unquestionably in the #1 spot for EVERY STAT (except Points Against, where Loveless managed to win out). Finishing 3rd wasn't enough to dethrone the reigning Tournament Champions.

For overall tournament record, I wasn't sure how to decide the tie between Grizzlies and Loveless. There are a number of various tie-breakers I could have used, especially since they have played an unequal number of games, but I decided to just call them tied and be done with it.

Also, Mercs still suck.

[Bob]

Tournament Masta!

I honestly don't get the Mercs. They do crappy in the league, then stomp on people during the tournament. I'm not sure what happens. I think I just play them more seriously during tournaments, maybe? I do try to maximize their stamina for tournaments so I can play Gafgarion in basically every game, so maybe that's why. Which means they likely won't do as well this tournament since it goes longer. But we'll see.

Mercs rule!

*car hits banana peel and drives off ledge*

[Bob]

I'm pretending to be busy. I mean, like, I'm REALLY busy, but also really easily distracted. And so, I made a spreadsheet tracking league and tournament wins. Then, I gave a point for bronze, two for silver, and three for gold, just to see by medals which team is the best. Currently, Grizzlies are on top because of their two golds, but if Loveless takes the current tournament, then they'll be tied with them.

If interested: forum.goatpen.org/Blitzball/BlitzStandings.xlsx

[Fleck]

For your viewing pleasure, if give you MORE STATS!

This is a chart showing how our respective teams have fared against each other in League Play so far.

Explanations:

The chart assumes that the team in the left-most column is the primary team, and that the teams listed along the top are secondary. So it should be read as "___" record against "____" with the top-row teams in the second blank. So the Goers record against the Aurochs is 3-0, while the Aurochs record against the Goers is 0-3. Long story short... don't try to read from the top down.

For funsies, I decided to color in the teams that each team does best/worst against (again, reading from the left, not from the top). Doing so has revealed some interesting truths about the League thus far. A team's "Worst Enemy" (colored in red) is the team that you have done the worst against. "Best Friend" is the team you've done the best against. In instances where multiple teams have the same record for best or worst, I give the award to both, rather than resorting to tie-breakers. Also, the Goers and the Aurochs have both been excluded from the best/worst categories, since almost all teams do the best against the Aurochs and the worst against the Goers.

Let me know if you have any questions or find any mistakes. I'll put up a similar chart for Tournament Play after the 4th Tourney.

[Fleck]

MORE STATS!

Okay, this one is a little different. Rather than comparing teams versus teams, I have compared formations versus formations. What I found was, well, I wasn't surprised--because I expected results along these lines--but I was surprised by the extent of the results. Anyway, here's my (very sloppy) stats.


AGGREGATE FORMATION STATS BY SEASON[/u]


SEASON 1

NORMAL (1) v CENTER (4) ties: (3)
WIDE (4) v NORMAL (5) ties: (0)
1-3-0
CENTER (4) v WIDE (0) ties: (0)
1-0-0

SEASON 2

NORMAL (7) v CENTER (8) ties: (4)
WIDE (1) v NORMAL (2) ties: (4)
CENTER (2) v WIDE (1) ties: (0)

SEASON 3

NORMAL (1) v CENTER (3) ties: (2)
WIDE (0) v NORMAL (3) ties: (0)
CENTER (2) v WIDE (0) ties: (0)
CENTER (3) v JECHT (1) ties: (1)
CRESCENT (0) v SWEEPL (1) ties (0)
CRESCENT (0) v SWEEPR (1) ties (0)
CRESCENT (0) v NORMAL (1) ties (1)
RELAY (0) v CENTER (1) ties (1)
NORMAL (2) v JECHT (2) ties: (1)
JECHT (1) v SWEEPL (0) ties (0)
RELAY (1) v SWEEPL (0) ties (0)
NORMAL (1) v SWEEPL (0) ties (0)
JECHT (1) v CRASHL (0) ties (0)
NORMAL (0) v SWEEPR (1) ties (0)
RELAY (2) v NORMAL (0) ties (0)
JECHT (0) WIDE (1) ties (0)
NORMAL (0) v CRASHL (1) ties (0)
WIDE (2) v CRASHR (0) ties (0)

SEASON 4

NORMAL (0) v CENTER (8) ties: (2)
WIDE (3) v NORMAL (0) ties: (0)
CENTER (2) v WIDE (1) ties: (1)
CENTER (1) v JECHT (0) ties: (0)
RELAY (1) v CENTER (0) ties (1)
NORMAL (1) v JECHT (1) ties: (1)
NORMAL (0) v SWEEPL (2) ties (0)
RELAY (3) v NORMAL (2) ties (0)
JECHT (1) WIDE (0) ties (0)
RELAY (0) v WIDE (2) ties (0)
CENTER (1) v CRASHL (0) ties (0)

TOURNAMENT 1

NORMAL (2) v CENTER (2)

TOURNAMENT 2

NORMAL (1) v CENTER (1)
WIDE (1) v NORMAL (0)
CENTER (2) v WIDE (0)

TOURNAMENT 3

NORMAL (0) v CENTER (0)
CENTER (1) v WIDE (1)
SWEEPL (0) v CENTER (1)
CENTER (2) v JECHT (0)
NORMAL (0) v RELAY (1)
CRESCENT (0) v JECHT (1)
NORMAL (1) v JECHT (0)

TOURNAMENT 4

NORMAL (0) v CENTER (2)
CENTER (2) v WIDE (0)
SWEEPL (0) v JECHT (1)
CENTER (3) v RELAY (0)
WIDE (1) v JECHT (0)

Commentary:

Those numbers beneath some of the stats in season one are the results of the resim matches that Bob did to correct the glitches in Juice's games. I left them out of my figuring because they didn't affect much, but I included them here in case anyone was curious. And yeah, sorry that I don't have my things consistently on one side or another. This turned out to be a bigger project than I thought (so far almost 2 hours have gone into this) and I skimped on cleaning up the stats for the sake of getting this out sooner. I'd probably have gotten this out quicker if y'all didn't have an orgy of variations in Season 3. Honestly, that took up most of my time. You guys and your fancy formations!

Also note that any time the same formation encountered itself (normal v normal, center v center, etc.) it was excluded for being irrelevant. Hence the relatively low numbers.

Next we have the more telling stats:


AGGREGATE FORMATION STATS BY FORMATION[/u]


NORMAL (30-50-17)

NORMAL (12) v CENTER (28) ties: (11)
WIDE (9) v NORMAL (10) ties: (4)
NORMAL (4) v JECHT (3) ties: (2)
RELAY (6) v NORMAL (2)
NORMAL (1) v SWEEPL (2)
NORMAL (0) v SWEEPR (1)
NORMAL (0) v CRASHL (1)
CRESCENT (0) v NORMAL (1)

CENTER (55-17-15)

NORMAL (12) v CENTER (28) ties: (11)
CENTER (15) v WIDE (3) ties: (1)
CENTER (6) v JECHT (1) ties: (1)
RELAY (1) v CENTER (4) ties (2)
CENTER (1) v CRASHL (0)
SWEEPL (0) v CENTER (1)

WIDE (18-26-5)

WIDE (9) v NORMAL (10) ties: (4)
CENTER (15) v WIDE (3) ties: (1)
JECHT (1) v WIDE (2)
WIDE (2) v CRASHR (0)
RELAY (0) v WIDE (2)

JECHT (9-12-3)

CENTER (6) v JECHT (1) ties: (1)
NORMAL (4) v JECHT (3) ties: (2)
JECHT (1) v WIDE (2)
JECHT (2) v SWEEPL (0)
JECHT (1) v CRASHL (0)
CRESCENT (0) v JECHT (1)

RELAY (8-8-2)

RELAY (1) v CENTER (4) ties (2)
RELAY (6) v NORMAL (2)
RELAY (0) v WIDE (2)
RELAY (1) v SWEEPL (0) ties (0)


Commentary:

So yeah, obviously, the numbers next to the name of each formation is that formations overall win-loss-tie record. And it slants MASSIVELY in favor of the center formation. The only formation that seems to be able to threaten Center is the Normal formation, and even it can't even quite reach half of Center's wins.

Does this mean that Center is OP? Maybe, but this doesn't prove anything. There's a lot to consider. For one, a lot of awesome teams (Loveless, for instance) tend to favor the Center formation. Is center great and therefore Loveless is great? Or is Loveless' greatness influencing Center? Hard to say. Conversely, the Guests Who favor the Normal formation and tend to lose a lot, so that skews data as well.

Also, the note that the Goers and the Aurochs both use Normal. Aurochs almost always lose and the Goers don't win enough to offset the Auroch's losing, which skews data in favor of Center. In addition, these numbers take tournaments into account, and neither the Goers nor the Aurochs are there to represent Normal. So it's hard to say how much of an advantage the Center formation has.

It is clear that specialty formations bounce off Center like rain off a tin roof, but both Jecht and Relay do quite well against Normal. There isn't enough data to say much about how well they do against Wide, but so far Wide has the advantage.

I only did averages for the five most common formations (normal, center, wide, Jecht, and relay). If you want to know the stats for the crash, sweep, and crescent formations, DO 'EM YERSELF! I'm too lazy.

Thoughts?

[Bob]

Until this season, Grizzlies were also almost exclusive Center formation. But Grizzlies are still winning a lot this season and I'm varying my formations now, so I do think it's skewed by GB and LO. That being said, Center is a good formation for most teams because you can make up for a weaker defense with Center. Teams like the Mercenaries don't do amazing with Center (they do okay though) because they have a lot of defenders that don't need to tag-team somebody to take 'em down. So it does still greatly depend on your team dynamic.

Also, SM7 is starting to do well now that they're using Center more often and not exclusive JECHT. So yeah, I think Center is pretty OP, but again, gotta have the right stats, otherwise I think there are better options depending on the situation. So I think formations are doing their jobs and allowing variation in dynamic. Yes, Center is good, but probably not good enough that all teams should swap over to it. Maybe we should make a formation that is completely geared toward countering the Center formation though...hmm... Not sure what that'd be though.

[countlieberkuhn]

You'd think that the counter to centre would be wide, as in theory that formation would be able to bypass the bulk of centre's power by simply going avoiding the majority of it. However even MAG, who have the passing power to make the wide formation more viable don't seem to perform overly well. Although I'm beginning to think of a possible way to improve upon it, as far as dealing with centre goes.


---------------LF

-----LD


G---------------------------RF


-----RD

---------------C


Something like that could work so long as your LF and C had superior passing stats, perhaps.

Or even this:


----------------------------------LF


G

----C


---------------LD


--------------------------RD


Where the RD has an exceptional passing ability, and by constantly passing forward until you hit the RD player, you'd draw centre's aggro to one side of the field (which may be easier to do since they're closer together), and then do a long pass to the LF who would have a clear shot. It all depends on whether this theory would translate well in the players' AI though.

[Bob]

Your bottom illustration is missing RF (or perhaps missing RD and mislabeled RF?), so I'm a bit confused. BUT, it makes me think that we should take one of the left/right formations (Crash or Sweep) and adjust it so that one side is doing something more Center-like and the other more Wide. Something like:

----------------------------------------LF

--------------LD

--------------------C

--------RD

---------------------------RF

I feel like the problem with both CrashLR and SweepLR is that they really weaken your defense, but this might allow for a still solid defense, but allow you to use a good passer up front to say, get the ball to the LF for a close shot or shoot from a distance from RF.

Wouldn't do this 'til next league though, and only if others like the idea, but throwin' it out there.

[Fleck]

I think it's the wrong approach to try to look at it as "how can we build a formation that can beat Center?" Or rather, we CAN look at it that way, but we should really be looking at making two new formations: one that can counter Center and one that performs, on average, as well as Center. Because if we just make a formation that counters Center, then we run the risk of ending up in a Pokemon "fire beats plant, plant beats water, water beats fire" situation where guessing your opponent's formation is 90% of the game.

Ideally, we could make formations that are tailored to particular play styles, but we've already tried that and haven't gotten the greatest results from it.

Basically what I'm saying is that I have no idea what the answer to the problem is.

[Bob]

I do agree that we shouldn't necessarily look at how to beat Center. We don't want Rock/Paper/Scissors, we just want adjustments based on play style.

That's why my proposal was more to find another formation that might have similar success to Center. I like the idea of playing one side more forward than another because it'd take the proven Center format, but put one of the forwards in a better striker position. Reason this would be different from Center is that players try to pass as far forward as they can, so as long as a team has a good enough pass stat, they will pass to one striker a lot more frequently than the other. Could be good for teams with one exceptional striker like Mercs (Gafgarion), Guests (Larsa), SM7 (Jecht), Marlboros (Tidus), and so forth.

So maybe my wording about countering Center was off, because I agree that we're not looking to obliterate Center, just create other formations that are equally awesome.

Also: keep up the analysis, guys (mostly Fleck, but I know Wyvern's going to do things, too)! I love the stats about teams, formations, and all the variables. It's nice to know what is and isn't balanced about my program so I can continue to improve it...and so I can make my teams better than all of yours

[Fleck]

Speaking of that, I'd like to know what Gabe is planning to do so that we don't end up wasting a bunch of time copying each others stats.

For me, my next thing was to make a Plus/Minus stat for each individual player (not individual goals scored; that would take too long. Or at least I'll save that for later). Basically a Plus/Minus stat means how many goals are scored when Player X is on the field versus how many goals are scored against. It's a thing I picked up from hockey (although other sports use it too, I think).

It should be useful for giving everyone a rough estimate of how effective each player is (even defenders). Players who have negative stats are letting through too many goals and players that are positive are helping the team win. I don't expect the results to be too surprising, as I think everyone knows who's helping and who's hurting their teams, but it'll be nice to be able to bust out a chart and point fingers.

[wyvernxk7]

I'm just doing a master chart of all games played by each team, accounting for overall wins, losses, and ties, and highlighting each game accordingly. It's kinda simple compared to the cooler charts you're doing, but I'll find it interesting...

[countlieberkuhn]

I'm actually not too clued up on who hurts my teams, I tend not to notice trends too much in the text files. The only person I really spotted it on was old Balthier, who is now happily remade and more useful!

Also, what about player remakes affecting the plus/minus stats? That could be a problem if you're basing this data over several leagues, although it should be fine on a league by league basis.