AI batting - innings pacing experiences

[Jimmytheshoe]

I am totally agree with Ashley . AI batting must be improved in t20 games. In t20 cricket average score is around 160 in real world that never happened when AI bat. At pro level AI score never cross 130 against 2 star rating bowling attack. And When they chase 140 they never get it . In last 5 over they don't attack while they have 6 or more wickets in hand. They also miss most of free hits . So please in next update AI batting must be improved specially in shorter format.

Sorry for my bad English.

Ashley is bang on AI are really good accept for the last few overs. On quite a few occasions I have had the AI batting second with less than 15 runs to get and 2 or so overs to go and they have blocked out at least half of the deliveries with no attempt to even try and get the runs required.

 

[adikal]

Completely agree with Ashley on the AI pacing issues in T20s.

I mainly play T20s and on Legend difficulty. I've seen issues with the AI pacing not just towards the end of the innings, where they aren't aggressive enough even with wickets in hand (if the game had a scoring histogram of runs-per-over, it would look quite unrealistic towards the slog overs), but crucially, also when chasing targets. In particular, when chasing a small target (<100), the AI openers would play unnecessarily aggressively in the first few overs. Even if its understandable that they would want to take advantage of the fielding restrictions, what was surprising was they continued to do so even after a wicket or two fell -- the new batsmen were equally aggressive, which shows a lack of match situation awareness. Only after they lost 5 wickets or so did they start playing a little more defensively and got me back into the game, which I eventually won.

I've actually stopped playing for a while because of the AI pacing. I don't get time to play longer formats and I need a real challenge for T20s. And I don't like simulating the AI scores as I quite enjoy the thrill of bowling, setting fields etc.

 

[Gamer Pradosh]

Completely agree with Ashley on the AI pacing issues in T20s.

I mainly play T20s and on Legend difficulty. I've seen issues with the AI pacing not just towards the end of the innings, where they aren't aggressive enough even with wickets in hand (if the game had a scoring histogram of runs-per-over, it would look quite unrealistic towards the slog overs), but crucially, also when chasing targets. In particular, when chasing a small target (<100), the AI openers would play unnecessarily aggressively in the first few overs. Even if its understandable that they would want to take advantage of the fielding restrictions, what was surprising was they continued to do so even after a wicket or two fell -- the new batsmen were equally aggressive, which shows a lack of match situation awareness. Only after they lost 5 wickets or so did they start playing a little more defensively and got me back into the game, which I eventually won.

I've actually stopped playing for a while because of the AI pacing. I don't get time to play longer formats and I need a real challenge for T20s. And I don't like simulating the AI scores as I quite enjoy the thrill of bowling, setting fields etc.

Yeah they don't have "slog every ball" attitude when they have wickets in hand and its those 15-20 overs...Very odd...I can even bowl a maiden in the final over [in career]....

 

[vishaalmmm]

But realistically speaking, you do see players blocking out maiden overs in a t20 game. What scores on average have you seen AI score in a t20 and at what difficulty? I usually find myself chasing 140+ scores in a t20 game on legend, but maybe either I'm bowling reasonablly well or a little on the horrid side.

Yes , you do see players blocking out maiden overs in t20 game but not in slog over.So definitely AI batting is not up to the mark in shorter format . They don't play according to the situation. 160-170 is average score in t20 but they never score 140+ at pro level or legend level. They are not able to chase down 140+ in 20 over while they have wickets in hand. I have played all cricket games like blc99, blc05, blc07, ashes09, ea07. AI batting is not up to the mark in all cricket games. I hope big ant studio consider this and will improve AI batting in next update.

 

[superfreddie]

Have to agree with the OP.

Also I was playing in career mode the other day and my team needed to win the last group game to get to T20 finals day. I got out with a couple of overs to go and Jos Buttler needed 11 off the last over. Good chance of that I thought so I decided to spectate the last over, Buttler got 3 and didn't look to smash any of the last over deliveries, I should probably have simulated.

 

[adikal]

Yes , you do see players blocking out maiden overs in t20 game but not in slog over.So definitely AI batting is not up to the mark in shorter format . They don't play according to the situation. 160-170 is average score in t20 but they never score 140+ at pro level or legend level. They are not able to chase down 140+ in 20 over while they have wickets in hand. I have played all cricket games like blc99, blc05, blc07, ashes09, ea07. AI batting is not up to the mark in all cricket games. I hope big ant studio consider this and will improve AI batting in next update.

Yes, to be fair to BigAnt, you are spot on. AI is usually the last thing that cricket devs pay attention to (or for that matter most gaming devs), and this particular AI problem -- AI pacing during a cricket match -- is particularly non-trivial. Still, hoping to see some breakthrough in the next version

 

[AshleyWilliam]

AI in ODI's look to score at around 5 rpo after the powerplay overs and hit out at the end of innings = Perfect.

AI in Test Matches look to score at around 3 rpo = Perfect.

AI in T20 look to score at around 4 rpo after the powerplay overs. Not perfect. Change this 4 to a 7 or 8 = perfect.

 

[AshleyWilliam]

I reckon we should wait for other people's opinions before considering making some changes.

Please see above!

 

[Ren]

I would love an AI setting in the pause menu where we could set whether batsmen not controlled by the player are attacking, opportunist or defensive.

I've had some 4 day career matches where we need to bat for a draw and the batters at the other end keep edging on attacking shots or holing out in the deep.

I feel that t20 AI is slightly out of balance too, but I'd also hate to see them getting all out for 100 in 10 overs.

 

[79cricket]

I must say I have never seen AI bowl or bat like in this game, as I play my career mode enjoying watching my AI batsmen bat.

For this T20 problem to be honest not seen anything to bad yet, the only thing is how low the AI edges the ball vs Me or my AI partners. something is wrong there.

 

[eaglesrjh]

@fiction

got to agree that there seems to be a pacing issue with T20

background, i'm having a bit of fun and playing on Rookie, mainly as a bowler so not getting to bat at all.
I simulate to career player at all times.
these stats from Indian T20 league but have noticed similar in other T20 matches

played nine matches and here are some stats.

batting first, without me batting to artifically change it we average 10.2 runs per over, pretty good.

however batting 2nd
we average. 5.1 per over, that's half the first innings rate.

now dont get me wrong we have chased some lowish scores so can understand the run rate being a bit lower but we have failed to get some despite having plenty of wickets in hand or struggled to get lower scores.

for example,
chasing 131, we where 94/2 after 20 overs, so 4.7 per over despite plenty of wickets in hand
took 18.2 overs to chase 85 (4.74 per over)

101/4 chasing 143, run rate (5.05) again despite wickets in hand.
107/3 chasing 120 run rate (5.35) again plenty of wickets in hand.

if you look at the real world scores then the expectation would have been to comfortably win all these matches or at least lose more wickets trying to chase them. 94/2 chasing 131 is not realistic in my eyes, a couple of players at least would have had a real go at upping the rate and either succeeded and won the match or got out trying, if it was 94/8 then i'd probably not be too worried about it but 94/2 just says the players could not be bothered chasing the total!