I've been doing that with ODIs, will probably add a rolling window for T20s too.With each passing season, T20 par scores are clearly rising. When T20 started, the par was around 150. But nowadays par is possibly somewhere between 170 and 175. Do you take that into account?
It's important. Your current odds are biased towards the team batting first. When the team batting first is scoring around 160, your odds are assigning around 50% winning probability to them - but it should be much less than that as you can see the par score in IPL matches this season seems to be around 175 (180+ when playing in Mumbai and Bangalore).I've been doing that with ODIs, will probably add a rolling window for T20s too.
I feel like an idiot asking this but wanted to confirm - there is no powerplay or anything similar in T20s right?It's important. Your current odds are biased towards the team batting first. When the team batting first is scoring around 160, your odds are assigning around 50% winning probability to them - but it should be much less than that as you can see the par score in IPL matches this season seems to be around 175 (180+ when playing in Mumbai and Bangalore).
Just the first six overs one.I feel like an idiot asking this but wanted to confirm - there is no powerplay or anything similar in T20s right?
Actually I checked my FT20 db (IPL, BBL, CPL), and the average first innings score has been remarkably steady since the beginning. The only discernible difference I see is pre-2014 and post, where pre-2014 the average score is 156 and post it is 160 - just a 3% difference.It's important. Your current odds are biased towards the team batting first. When the team batting first is scoring around 160, your odds are assigning around 50% winning probability to them - but it should be much less than that as you can see the par score in IPL matches this season seems to be around 175 (180+ when playing in Mumbai and Bangalore).
Some kind of factor will work.Actually I checked my FT20 db (IPL, BBL, CPL), and the average first innings score has been remarkably steady since the beginning. The only discernible difference I see is pre-2014 and post, where pre-2014 the average score is 156 and post it is 160 - just a 3% difference.
I do see a clear difference for the last year's FT20s (04/16 - now) though where the average score goes up to 165 - this might be what you're pointing out. So it's a very recent phenomenon. Would probably just have to add a ad-hoc adjustment because ignoring all pre-04/2016 matches for odds is not an option (leaves <100 matches to work with).
Added an adjustment for scores post-2014 to reflect how T20s are more high scoring now.Some kind of factor will work.
For example, if pre-2014 average was 150 and the current average is 165 then for pre-2014 matches for all scenarios include a (165-150)/150 i.e. 10% factor. Chasing 80 in last 10 overs with 7 wickets remaining will be like chasing 88 in last 10 overs with 7 wickets remaining now.
Bug in code - wasn't around to fix it so script wasn't running. Early days.Just wondering why 3 overs have been bowled and yet the KXIP vs GL match isn't showing in live odds...
same bug today for the KKR vs RPS match?Bug in code - wasn't around to fix it so script wasn't running. Early days.
Unfortunately I'm at work so can't look into atm :/same bug today for the KKR vs RPS match?
Unfortunately I have to catch these issues live on a day I'm working from home.Northants vs Warwickshire has frozen.
One of these days I'll be around to catch these.Why did the SRH odds suddenly dipped close to zero for a moment there?