Originally posted by ress5609
Originally posted by atingwald
so you are saying that he built this random generating program for all these teams, but he couldn't do it for say 100 teams of his own... with special builds built in to each of those 100 teams??
Those 100 teams could play each other over and over on a random loop, and the statistical anomalies could not be determined thru statistical analysis??? So a computer is slower than the humans?? LOL
I find that very hard to believe.... in 100 teams, you could generate millions upon millions of matchups, and the oddities would expose themselves in two weeks, guaranteed.... every two weeks, you can adjust those 100 teams, and that would lead to new tests...
how long do you think a game truly lasts??? with a random generator, it takes minutes at most... you can run millions of games in two weeks....
I think you misunderstand what the "easier said than done part" is. You're assuming that all bugs can be found through statistical analysis. Unfortunately, that is not the case. There's no guarantee that a particular bug will show up as a statistical anomaly, and further more writing the logic in code to detect any possible bug would take longer than just sitting down and watching every single play of every single sim.
team A plays team B 100 times... results should be pretty similar over 100 games...
team C plays team B 100 times.. again, these results should be pretty similar over those 100 games
team A has guys with 20 points more confidence than team C... wouldn't you be able to see a statistical difference??
you don't have to write that into code... you can print out data and see the % difference....
Originally posted by atingwald
so you are saying that he built this random generating program for all these teams, but he couldn't do it for say 100 teams of his own... with special builds built in to each of those 100 teams??
Those 100 teams could play each other over and over on a random loop, and the statistical anomalies could not be determined thru statistical analysis??? So a computer is slower than the humans?? LOL
I find that very hard to believe.... in 100 teams, you could generate millions upon millions of matchups, and the oddities would expose themselves in two weeks, guaranteed.... every two weeks, you can adjust those 100 teams, and that would lead to new tests...
how long do you think a game truly lasts??? with a random generator, it takes minutes at most... you can run millions of games in two weeks....
I think you misunderstand what the "easier said than done part" is. You're assuming that all bugs can be found through statistical analysis. Unfortunately, that is not the case. There's no guarantee that a particular bug will show up as a statistical anomaly, and further more writing the logic in code to detect any possible bug would take longer than just sitting down and watching every single play of every single sim.
team A plays team B 100 times... results should be pretty similar over 100 games...
team C plays team B 100 times.. again, these results should be pretty similar over those 100 games
team A has guys with 20 points more confidence than team C... wouldn't you be able to see a statistical difference??
you don't have to write that into code... you can print out data and see the % difference....






























