IF THEN ELSE, INC.
Pro sports prediction programs for the
Windows 9.x/NT/XP/Pocket PC Operating Systems

'...could betwo of the most important programs for the sports gaming public yet.'

'..are invaluable in providing possible scores of major professional sporting games

'... Great Simulations... do the almost impossible and they do it
       very effectively.'


| Introduction | Features | Screenshots | Availability | System Requirements | 
| Recommended | | Download and Purchasing Links | Useful Links | 

Introduction :

FUBAWCE & FUHBAWPPC

Just say it with a real drawl ... this heah game is fuhbaw and this one heah is the boss. FuhBawCE/FuhbawPPC predicts the results of USA professional gridiron football league games using published up to date statistics - Offense and Defense rushing, passing, penalties, Kicking, fumbles, sacks, interceptions and team home and road results. All stats are run through a game playing engine with some variable constants to calibrate to actual results. Note, the results do have some randomness and are not repeatable. During the 2003 season, FuhbawCE correctly predicted 162 games versus 94 errors - a success rate of 62.5%! In the Usatoday.com "Pigskin Pool", the best results picker out of 14,148 contestants was at 184-72. FuhbawCE was in the top 9.5% of all players, better than 12,811 participants! On Over/Under picking FuhbawCE was 98-77-3 for 56%!

The program has options for either single game play-by-play mode or a rapid 8-game series with scoring averaged. The 8-game mode is used for game prediction. No guarantees of any kind are made regarding the performance of this program.

Included in the database are seven previous championshiop teams - ranging from the Miami team of 72 to the Tampa team of 02. these can be used compare present teams with the past. Note, in calibrating past teams it is assumed that average teams remain the same over the years - an "average 1978 team" would match and "average 2003 team".

HAAKEE

For NHL Major professional hockey league games. HAAKEE predicts the results of North American major professional hockey league games using published up to date statistics for individual scoring, overall team offense home and road, team powerplays home and road, team shorthanded home and road, team penalties home and road. All stats are crunched into a probability algorithm with some variable constants to calibrate to actual results. Note, the results do have some randomness associated and are not repeatable.

The program will run a series of 11 games between the two selected teams and provide an average score. A "Money Line" result will show the expected winner's advantage based on number of total wins out of 11. Note, in some cases the team with the lower average score may have the Money Line advantage. User could compare the HAAKEE line with published Sports Gaming Industry "Lines".

In considering all Ties as "Pushes" the testing version of HAAKEE performed at 59.2% accuracy comparing with actual game results from 05 Oct 00 - 12 Mar 01. And 57.0% accuracy from Oct 01 - 28 Mar 02. In 2003/4 from October to end-Mar HAAKEE picked 55.2% ATS winners out of 359 picks. No guarantees of any kind are made regarding the performance of this program.

The database contains user-update-able stats for all players and team home/road goal scoring figure, which are then calculated to calibrate against actual data.

BEISBOLL

BEISBOLL predicts the results of North American major professional baseball league games using published up to date statistics for team hitting, overall team offense home and road, team defense home and road and team pitching. All stats are crunched into a probability algorithm with some variable constants to calibrate to actual results. Note, the results do have some randomness associated and are not repeatable.

The program will run a series of 11 games between the two selected teams and provide an average score. A "Money Line" result will show the expected winner's advantage based on number of total wins out of 11 . Note, in some cases the team with the lower average score may have the Money Line advantage. User could compare the BEISBOLL line with published Sports Gaming Industry "Lines".

The testing version of BEISBOLL performed at 58% accuracy in 1999 and 53% in 2001 comparing with actual game results. 3.17% ROI is what an investor COULD have earned using BEISBOLL the baseball predicition program. Howzzat you ask? Well, here's how it worked out, in retrospect:

1 - Beisboll predictions were posted daily, as a matter of public record, throughout the 2001 season, mainly to the rec.gambling.sport newsgroup. Check it out on http://www.google.com

2. Final, end of regular season stats showed "wins", or correctly predicted games, against "losses", for a net 53% winning percentage.

3. Througout most of the season games were pseudo-wagered using $100 per game and the following rule:(a) bet on all underdogs, (b)only favourites costing less than -$165, and (c) only if Beisboll's winning percentage over the previous 5 days was >55%. However, in June, during a development experiment with the programming, all games were "bet" despite the record of the past five days and even for -$300 favourites. Unfortunately, Beisboll lost heavily that month, and on the whole season ended up as -$ . In mid-July the trials reverted to following the rules and the latter two months were quite successful. (31 July was exceptional bloodbath losing -$1305 on trade deadline day, probably should give that day a pass next year). In conjunction with these three rules, team batting records and lineups had to be updated regularly, one team per day was updated or if a major injury/roster change occurred.

4. Now, going back and "re-doing" the season with strict adherence to the 3 rules as stated above, Beisboll ends up winning $775 on the year, a 1.1% ROI, and an average take-home of $11 on only 72 days placing a "bet".

5. Modifying the rule (c) to only wager when the last five days win .pct was 60%, a player would have earned $1019, a 3.17% ROI with an average $30.88 daily take-home in only 33 days placing approx ten - $100 "bets".

6. So, I think it shows that this kind of probabilistic, programmatic baseball wagering can provide a positive return on investment. No guarantees of any kind are made regarding the performance of this program.

Users may amend or change the league data files to suit. Teams are set based on the final 2001 individual statistics late Marc 02 probable lineups, normalized to ~ 5000 At Bats. Earlier versions used 9 man lineups but it was found to be too difficult to to follow actual team lineups on a daily basis. Thus the nine "every day" starters on any team are averaged out together, with 20% more weighting to the top 5 batters.

For optimum accuracy variable constants and hard-coded team statistics must be updated and calibrated regularly. Contact author for Registration/Updates (see below).


Current Features :

SIMPLE, EASY TO USE INTERFACE

Note: USER (WITH INTERMEDIATE COMPUTING SKILLS) CAN EASILY MODIFY THE STATISTICAL DATA FILES TO KEEP THE PROGRAM "CURRENT".
VERY LITTLE RAM OR ROM MEMORY USE