I simulated 1000 England vs Ireland Test Matches - what do you think of the results?

[Anku31]

I’ve been building a cricket management game and one of my match-engine tests is to simulate 1,000 matches between two teams and look at the aggregate results.

The screenshot shows what the engine currently produces for England vs Ireland on a neutral pitch.

Ireland was able to beat England only 42 times out of 1000 - that's a 4.2% win ratio.

I’m curious how believable these numbers look to people compared with real world at this moment? Is 4.2% win rate for Ireland good enough? Or too low?

I’m particularly interested in the win/draw ratio and scoring numbers. I provided England vs Ireland example as it's a good test with two teams at the extreme ends of the spectrum.
Ignore the powerplay RR.. it was meant for the ODI/T20 part of the engine.

I’m building this match engine for Cricket Team Manager. If interested, the game thread is here - https://www.cricketweb.net/forum/th...ager-releasing-on-steam-october-1-2026.98647/

 

[Abhay423]

Depends on strength of the respective team and conditions. So recently before the first Ban vs Aus test, Ban were 40/1 to win the test according to the betting markets(around 2.5%), they won that and then were 10/1 to win second test. Now Ireland will be more competitive in England due to familiarity with conditions, so it will be much closer to the second number. Personally I would say around 7-8%. But the major issue with your model is the draw probability. Draws are never this short in this day and age unless there is extreme bad weather. Can't see 1 in every 4 and 1/2 game to be a draw.

I would price it like this:
England:83.5%
Draw:9%
Ireland:7.5%

The dismissal probabilities are mostly fine and in line with historical stats though stumping at 3.51% looks a bit high for English conditions, probably be around 1.5(not 100% sure) and can add extra 2% to bowled

Innings run line plus the win by innings are pretty accurate.

Overall everything looks good apart from draw %. If you adjust that, underdogs win prob would automatically go up.

 

[ataraxia]

View attachment 55445

I’ve been building a cricket management game and one of my match-engine tests is to simulate 1,000 matches between two teams and look at the aggregate results.

The screenshot shows what the engine currently produces for England vs Ireland on a neutral pitch.

Ireland was able to beat England only 42 times out of 1000 - that's a 4.2% win ratio.

I’m curious how believable these numbers look to people compared with real world at this moment? Is 4.2% win rate for Ireland good enough? Or too low?

I’m particularly interested in the win/draw ratio and scoring numbers. I provided England vs Ireland example as it's a good test with two teams at the extreme ends of the spectrum.
Ignore the powerplay RR.. it was meant for the ODI/T20 part of the engine.

I’m building this match engine for Cricket Team Manager. If interested, the game thread is here - https://www.cricketweb.net/forum/th...ager-releasing-on-steam-october-1-2026.98647/

73.8% with declarations seems high to me. Only one of the five English tests this season had a declaration (though the Lord's tests were very low-scoring). How many overs are there per day and when do typical first-innings declarations occur?

 

[Anku31]

Depends on strength of the respective team and conditions. So recently before the first Ban vs Aus test, Ban were 40/1 to win the test according to the betting markets(around 2.5%), they won that and then were 10/1 to win second test. Now Ireland will be more competitive in England due to familiarity with conditions, so it will be much closer to the second number. Personally I would say around 7-8%. But the major issue with your model is the draw probability. Draws are never this short in this day and age unless there is extreme bad weather. Can't see 1 in every 4 and 1/2 game to be a draw.

I would price it like this:
England:83.5%
Draw:9%
Ireland:7.5%

The dismissal probabilities are mostly fine and in line with historical stats though stumping at 3.51% looks a bit high for English conditions, probably be around 1.5(not 100% sure) and can add extra 2% to bowled

Innings run line plus the win by innings are pretty accurate.

Overall everything looks good apart from draw %. If you adjust that, underdogs win prob would automatically go up.

Thanks for the detailed review. Points noted. Yes, some recent tweak in the engine caused the draws to go 17% from 10% (overall, not just ENG vs IRE). My aim for the engine is around 14%, which is what last 25 years of test cricket has given us.
Nice observation on stumpings - noted. I would have missed it. Important small detail.

 

[Anku31]

73.8% with declarations seems high to me. Only one of the five English tests this season had a declaration (though the Lord's tests were very low-scoring). How many overs are there per day and when do typical first-innings declarations occur?

Yes. I have noticed that there have been fewer draws in past 5 years or so actually. I will tweak the engine to go down a bit on draws. The overs per day are standard 90. Rain and bad light are also a feature of the game. So count that in too. Regarding when do typical first-innings declarations occur - I have built the engine to be smart of deciding that. It depends on how much time has passed and how much runs are on board and then the team decides based on that. Same goes for 3rd innings declarations - current lead and overs left in the match. Declare only if a favourable result (i.e. a win) is possible, otherwise continue batting. There have been massive recent improvements in the game engine, which I am shipping with the Beta 2 version. Draws is something I guess I need to re-tweak again for Beta 3.