Warrior General Games

Week 7: 14-2 (6-2 Alpha, 8-0 Zeta [did not post Zeta])
Week 8: 14-2 (8-0 Alpha, 6-2 Zeta)
Week 9: 13-3 (7-1 Alpha, 6-2 Zeta)
Week 10: 10-6 (6-2 Alpha, 4-4 Zeta)
Week 11: 13-3 (7-1 Alpha, 6-2 Zeta)
Week 12: 14-2 (8-0 Alpha, 6-2 Zeta)
Week 13: 11-5 (5-3 Alpha, 6-2 Zeta)
Week 14: 13-3 (7-1 Alpha, 6-2 Zeta)
Week 15: 11-4 (7-0 Alpha, 4-4 Zeta) [tie was thrown out]

FYI, numbers are confidence (basically a percentage chance of win).

Alpha vs. Zeta
--------------
74.35 - Hungary Hungry Hippos over Texas City Storm
63.50 - Brown Bay Packers over Green Bay Cheese Heads
58.92 - Romania Roadhouse Rowdies over Bergen Storm
56.38 - Rotterdam Rampage over Odessa Pipers
56.11 - Minnesota Warriors over Athens Gods
56.06 - Chernobyl CHUDS over Rammstein Rampage
55.88 - Macedonian Phalanx over Diaper Vipers
55.66 - River City Ransom over Nizhny Novgorod Bulldogs
54.94 - Moscow Red Army over Moscow Maniacs
54.13 - East Berlin Blitzkrieg over Warsaw Sabotage
53.84 - Minsk Fire over Crete Minotaurs
53.54 - Chernobyl Disaster over Boston Piggers
52.84 - Sparta Battle Chickens over Targoviste Impalers
51.98 - Krakow Krakodiles over Belgrade Bears
50.90 - Liverpool Poison over Barcelona Chupacabras
50.03 - The Dark Knights over Spanish Inquisition

Looks like 12-4 Alpha. Should be some good games.

Interesting, not what I expected...however most of the games are in the 50% area, so there should be some really good games!!!

liverpool needs to lose

Originally posted by allenc
50.03 - The Dark Knights over Spanish Inquisition



ummm....really?

56.11 - Minnesota Warriors over Athens Gods

Well... Don't quite see that happening, but not a shock that it was predicted that way.

I was surprised by the Dark Knights/Spanish Inquisition numbers as well. I don't even think it will be close. Inquisition should take it easily. Not surprised by Warriors/Gods numbers though. Should be another win for the Gods.

Originally posted by Nietzsche
56.11 - Minnesota Warriors over Athens Gods

Well... Don't quite see that happening, but not a shock that it was predicted that way.

Ha, true, you guys have had to garner your own respect all season!!! But the way you guys upset everyone, I'd take being the underdog in this situation...

Originally posted by mphil
liverpool needs to lose

+1

Not sure about those percentages. I know there is a formula and you didn't just make it up allenc, but they really need looking at.

74 percent for Hungary over Texas City? So 1 out of 4 times those team sim, Texas City will win? Highly unlikely.

That's not exactly what the percentage is saying. That's why I say "basically a percentage chance of win". Essentially, I get a strength number for each team based on items from both teams. Then divide the highest number by the sum of the two. So, it is possible to get close to 100% in the formula, but based on numbers produced by GLB teams, you will only see numbers between 50 and 90 or so. Highest I've seen is 91% I think. And I'll admit, the formula needs tweaking.

Originally posted by allenc
And I'll admit, the formula needs tweaking.

Maybe... You know, if you look at match up on the games you predict we would loose... we are normally 3-6 points behind in overall matchup as well. So... it may not be your formula that needs tweeked as much as the strength evauation factors on GLB.

Or maybe... some parts just cant be measured quite right
Or maybe... The Gods are just weird quirks -- LOL

Originally posted by Shardaddy
Originally posted by allenc

50.03 - The Dark Knights over Spanish Inquisition



ummm....really?


^this


needs tweaking

Originally posted by Shardaddy
Originally posted by Shardaddy

Originally posted by allenc


50.03 - The Dark Knights over Spanish Inquisition



ummm....really?


^this


needs tweaking

Well... there is only a 75% prediction rate in Zeta... So that (all by itself) gives you a 25% chance to win -- LOL.

But agreed... What is lacking in the formula is any trend analysis. However, I have rarely ever seen a purely statistical evaluation include actual trand analysis because they are so hard to program into the formula without over or under weighing its significance... That, and many number crunchers dont know how to do trend analysis because they over evaluate minute statistics rather than the bigger picture (no idea if that is the case in the numbers being used, but would be willing to guess that it is).

GG Today Liverpool

GG Warriors