Ordinal Logistic Regression

[dalkent]

I've been building models for a while and done some research and its seems most football models are either Rating or Poisson models or a combination of the two. I haven't seen anyway Ordinal (or Ordered) Logistic regression models, specifically for predicting 1X2 results. Has anybody experimented with this? I found it works betting that the other two approaches although that could simply be due to my particular models.

[Smoothplanet]

Re: Ordinal Logistic Regression http://www.r-bloggers.com/footbal-ordinal-model-examination-and-predictions/ talks about

[dalkent]

Re: Ordinal Logistic Regression

http://www.r-bloggers.com/footbal-ordinal-model-examination-and-predictions/ talks about

Thanks will take a look at that.

[paparainbow]

Re: Ordinal Logistic Regression any luck in this?

[dalkent]

Re: Ordinal Logistic Regression I have built the models for 1X2, under/over and correct score which i have converted to Asian handicap. As to weather they work, well if you had asked me a month ago i would have said working fantastically but after a bad Feb not so great. Only time will tell. For 2 weeks now though i have bet as a trial on almost every match using the asian handicap model. Both weeks there were about 40 matches and I broke even each time. So maybe its working and i should forget trying to select a few. I will keep trialling