[Bigredfyah]

I know this sounds stupid. What if I were to run simulations on sports games with up to date rosters add any injuries, weather to simulate sports games. For example let the patriots cpu play the eagles cpu 5 or 10 times and then add all the scores up and divide by either 5 or 10 then compare to the lines and see if theres any value. Am i out of my mind? What is the difference between doing this and all these simulation models out there. How is this any different? Has anyone thought about this or tried it. I only own Madden and wish I could try it out for football

[DwightShrute]

I know this sounds stupid. What if I were to run simulations on sports games with up to date rosters add any injuries, weather to simulate sports games. For example let the patriots cpu play the eagles cpu 5 or 10 times and then add all the scores up and divide by either 5 or 10 then compare to the lines and see if theres any value. Am i out of my mind? What is the difference between doing this and all these simulation models out there. How is this any different? Has anyone thought about this or tried it. I only own Madden and wish I could try it out for football

yes

[Bigredfyah]

Lol. Well it was just a thought . I'd like to try it for the hell of it. I don't see what the difference is between this and all the geeky simulation software.

[DwightShrute]

[VeggieDog]

Do it. What have you got to lose. Well, except money, of course. But you don't have to make actual bets while you're testing it out. Let me know what happens.

[Bigredfyah]

Wouldnt play for real money at first i only have Madden don't think there's enough time left in NBA or NHL to try this year. I'm not sure how mlb would work with the bullpen and everything could possibly do first 5 inning bets. I'll probably just wait until football starts again and bring this up again.

[BeatTheJerk]

I’d definitely be interested to know how it works out for you !

[KVB]

It's not the information, it's how you use it.

You need to run more than 5 or 10 simulations, use good software, then compare your numbers to the line to see if there is actual, long term value. You will have to run regressions but won't be able to simulate past performances.

The determination of value will be constantly subjective and float, it won't be very good.

But that doesn't mean you couldn't simulate a certain card and see if there is relevance in regards to the flow of money.

In the end, what you'll find, is that the simulations will be in line with common perceptions and ratings. After all, the most advanced games created ratings for each element.

Sometimes all the stats and numbers, and everyone's opinion, just line up. Those times will likely be displayed in your simulations.

I wouldn't be surprised to see the latest games simulate fairly close with those obvious exceptions that surprise everyone.

But will that make a profit in the market, not likely. You will need to pick your spots and likely use other information to make a good decision; buy, sell, or pass.

In the end, the other information will probably take care of the decision making, save the few exceptions where money is being stacked and taken down, which could be revealed with the simulations.

[A4K]

You need the ability to add weather as well as run thousands upon thousands of simulations to flatten out the extreme ends of the results. The more you can fill the bell curve the easier it will be to find a fair average outcome.

[Waterstpub87]

People tend to do this often with football using Madden.

I looked at using Maddon ratings on players prior to the last season. Maddon ratings weren't predictive of anything. I haven't played madden in close to 10 years, so I dont know if you can just do a massive simulation, but waiting for it to run 1,000 games is going to take a while.

Build a simulation is another program. I use excel to run 1,000 simulations for MLB, International Baseball, NCAAFB and NFL. Normally per game it takes:

20 Minutes MLB
10 Minutes International Baseball
30 Minutes for both NCAAFB and NFL.

Run overnight 2 days before to be able to bet the openers.

[A4K]

There used to be a home made website where the guy ran thousands of MLB simulations on every game. For the life of me I cannot find it.

[Bsims]

This sounds like a fun project, but I wouldn't count on getting useful results from this. Simulating sports games to get results that can be used for betting is a difficult project. I spent several years doing baseball simulations (hence the name Bsims) trying to get an edge before just giving up. Baseball is probably the best sport to simulate because it is pretty much a one on one game (batter versus pitcher) with lots of statistics available.


I have described my efforts in a post on my blog ole44bill.blogspot.com. Here are some of the issues faced.


· You need to select a simulation program.
· You need to run this against historical data to determine accuracy of results.
· You need to decide what player statistics to use,
· You must handle lineups and game strategies.
· You must run thousands of simulations for each game.
· You must develop a method of using these results to select wagers.

This is a very difficult task to try to accomplish. I could not succeed. I’m quite certain running Madden can’t be a reasonable approach. Good luck.

[TheMoneyShot]

I tried this with Madden 1998. Didn't work.

[Bigredfyah]

Thanks for the replies guys. I was gonna buy a baseball game for Xbox one but just realized they only make mlb the show for Sony. Pfffff

This doesn't mean much because the games have already been played and ratings adjusted but for the hell of it I simmed Vikings and saints 30 times. On average the Vikings scored 31.6 points and the saints scored 24.1 points I believe the spread closed at -5.5 For Vikings. I also simmed rams and falcon 30 times on average 29.7 for falcons and 24.0 for rams. I'm definetly going to be doing this for at least NFL next year and if results are good I'll put of cash on it.

[A4K]

This sounds like a fun project, but I wouldn't count on getting useful results from this. Simulating sports games to get results that can be used for betting is a difficult project. I spent several years doing baseball simulations (hence the name Bsims) trying to get an edge before just giving up. Baseball is probably the best sport to simulate because it is pretty much a one on one game (batter versus pitcher) with lots of statistics available.


I have described my efforts in a post on my blog ole44bill.blogspot.com. Here are some of the issues faced.


· You need to select a simulation program.
· You need to run this against historical data to determine accuracy of results.
· You need to decide what player statistics to use,
· You must handle lineups and game strategies.
· You must run thousands of simulations for each game.
· You must develop a method of using these results to select wagers.

This is a very difficult task to try to accomplish. I could not succeed. I’m quite certain running Madden can’t be a reasonable approach. Good luck.

I bookmarked your blog. I like where your head is at.

[phinfan27615]

flipping a coin would be easier and get you the same results

[A4K]

flipping a coin would be easier and get you the same results

Like we can afford a coin.

[captrobey]

I have always used this one for all of my picks