Warrior General Games

Thanks to all those that have volunteered to run these projections and post them for their league we now have over 70% of the leagues covered.

I'm still getting plenty of interest in the tool from others that just want to play with it. This was always in our plans, but the continued interest has spurred us to go ahead and release it tonight.

For an explanation of the purpose of this tool, see the thread below:

http://goallineblitz.com/game/forum_thread.pl?thread_id=447611

For other examples of what it can do, check just about any League forum (as I say, 70% or more have similar postings from this spreadsheet). If no one is posting it for your league, feel free to take that on. The main purpose here is to let people enjoy this and build a little positive buzz in the forums over the summer.

So, without further ado, here are the files:


For Excel 2007 w/ Formatting tool:
http://www.filefactory.com/file/4e57d4/n/LeagueProjectorTemplate1_1_xlsx
revised

For Excel 1997-2003 w/ Formatting tool:
http://www.filefactory.com/file/6baeeb/n/LeagueProjectorTemplate_xls

Formatting tool only (for those that got an early release)
http://www.filefactory.com/file/bca7b0/n/ProjectorFormat_xlsx


All these files are available free of charge. There is no need to register, or provide any information in order to download.

Thanks, this is neat, I think it works pretty well for predicting final records, but it needs some tweaking of the score predictions, they're pretty ridiculous sometimes, but maybe it's just because I'm in a league that has two terrible teams that skew the data.

I will use this perhaps... first see, very very complex!

nice work fellas, can't even comprehend what it took to do that.

Great Idea man. You could at lest offer people the opportunity to donate to you. You did a lot of work on this. I mean in the USA Pro league it went 16-0 and the scores where like right now.

Originally posted by brigmar
The spreadsheet went 16-0 for Week 7 with some very close results:
(Marauders,Saints,Superbowlers)

Intimidators 23-14 Globetrotters (23-3)
Marauders 63-0 Destroyers (65-0)
Dump Trucks 38-19 Daggers (32-10)
Bucks 8-21 Extreme (9-15)
Saints 25-18 Spartans (26-16)
Brute Squad 55-6 Just Go Krazy (48-0)
Bullies 10-17 Superbowlers (13-17)
Athletics 23-9 Supernovas (33-3)

Bulldogs 26-36 Skyhawks (13-26)
Mustangs 154-0 Gearheads (109-0)
Juggernauts 2-28 Pushovers (0-23)
Coyotes 115-3 Blazers (56-3)
Warriors 18-44 Nasty (35-10)
Warpath 42-3 Firefox (37-0)
Bullets 38-13 Sabercats (33-3)
Demon Lions 49-6 Wildcats (40-0)

Originally posted by GeneralMakaveli
Great Idea man. You could at lest offer people the opportunity to donate to you. You did a lot of work on this. I mean in the USA Pro league it went 16-0 and the scores where like right now.



I won't deny anyone the opportunity to donate to me, I promise.

wow, you were almost dead on, on quite a few of those.

great stuff, already did it for my league....got my concerned for a few games coming up...

Originally posted by kmart
Thanks, this is neat, I think it works pretty well for predicting final records, but it needs some tweaking of the score predictions, they're pretty ridiculous sometimes, but maybe it's just because I'm in a league that has two terrible teams that skew the data.

Yeah I ran it for other leagues, the bottom two teams in our league skew the score totals too much for an accurate prediction of ANY team. Those that have played them get many more points per game while those that have not don't have the added stats so a 17-14 game becomes 49-3.

Originally posted by caponesgun
Originally posted by kmart

Thanks, this is neat, I think it works pretty well for predicting final records, but it needs some tweaking of the score predictions, they're pretty ridiculous sometimes, but maybe it's just because I'm in a league that has two terrible teams that skew the data.

Yeah I ran it for other leagues, the bottom two teams in our league skew the score totals too much for an accurate prediction of ANY team. Those that have played them get many more points per game while those that have not don't have the added stats so a 17-14 game becomes 49-3.

I believe you have an error in the scores you entered. The algorithm is very capable of distinguishing between good and bad teams. It will not take two evenly matched teams and spit out a 49-3 score.

Check cells BK18:BL18 on the conference page. If you see numbers there, you've got errors in the scores you entered. If you see numbers in cells CY18:CZ18, there are errors in the scores for the other conference.

Originally posted by Satchmo-n-Dizzy
Originally posted by caponesgun

Originally posted by kmart


Thanks, this is neat, I think it works pretty well for predicting final records, but it needs some tweaking of the score predictions, they're pretty ridiculous sometimes, but maybe it's just because I'm in a league that has two terrible teams that skew the data.

Yeah I ran it for other leagues, the bottom two teams in our league skew the score totals too much for an accurate prediction of ANY team. Those that have played them get many more points per game while those that have not don't have the added stats so a 17-14 game becomes 49-3.

I believe you have an error in the scores you entered. The algorithm is very capable of distinguishing between good and bad teams. It will not take two evenly matched teams and spit out a 49-3 score.

Check cells BK18:BL18 on the conference page. If you see numbers there, you've got errors in the scores you entered. If you see numbers in cells CY18:CZ18, there are errors in the scores for the other conference.


I've been playing with this program for Africa Pro since the beginning of the season, and it is a ton of fun to play with. Most problems I encountered were due to inputting scores incorrectly, but Satchmo pointed it out, and I got it running like a dream, it is an amazing piece of work to get to play with.

so the game predications should be the same score for both teams right? right now i have one game that says 2-28 and the other team it says 2-20 for the same game..did i input something wrong?

Originally posted by TurboBoxer
so the game predications should be the same score for both teams right? right now i have one game that says 2-28 and the other team it says 2-20 for the same game..did i input something wrong?

If it is game 1-15, then yes, you have made an error. If it is a Gm 16, then not necessarily.

The algorithm cannot resolve more than about 60% of the games played in week 16 because there is no data for inter-conference play (no numbers for comparative analysis to crunch). In most cases where two good teams or two bad teams are matched up in Gm 16, the analysis will pick a different winner depending on perspective. That means it will likely expect both bad teams matched against one another to lose and both good teams matched up to win. You can use the average of the numbers to approximate some sort of score, but it will not have the accuracy of the other games. Averaging tempers all results toward the league mean, and is a poor substitute for individual data points.

When there is a consensus, the outcome is extremely likely and very obvious. Where there is not... well let's just say I wouldn't post a line.

yea its for the next game coming up....ill look into...can be anything else besides a incorrect score of a past game? (just double checked both conferences and everything looked right)

Originally posted by TurboBoxer
yea its for the next game coming up....ill look into...can be anything else besides a incorrect score of a past game? (just double checked both conferences and everything looked right)

If you have no numbers visible in cells BK18:BL18 or CY18:CZ18, then there is likely no error in data input. The next thing to check would be schedules. If there are visible numbers in row 18 below the scheduling cells, you have an error in that column. Or if the cell of the team name is still orange, you have an error in that row.

It is possible to have a scheduling error that is missed by the checker but, with cross-checking both directions, it's not very likely.

If you see none of problems described above, then send me a PM and I'll give an email to send the file. I should be able to check the file and send it back the same day.