grumbleguts
User name to post ratio checks out
Yep. One of the worst in world cricket. Can't stand the bloke and the way commentators jizz themselves over him. "Uncharacteristic turn over from Judd/Ablett" areas.
2023 ICC Men's Cricket World Cup Game 39 Australia v Afghanistan 7/11 1900hrs @ Wankhede Stadium
- Thread starter Gough
- Start date
- Tagged users None
Dogs_R_Us
Space Traveller
Even though he doesn’t like it himself.
Maxwell broke the algorithm
sherb
Hall of Famer
Not Found | Fox Sports Videos
www.foxsports.com.au
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
Reckon your list also warrants Symonds' 143 off 125 vs Pakistan coming in at 80 or 90/4, and Gilchrist's 149 off 104 in the 2007 Final.
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
I torrented it. Is worthy of 8GB on my hard drive rather than relying on Kayo from where one day it might simply vanish unaccountably.
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
To be honest I've grown a bit sceptical of these probabilities throughout the tournament. Consider this passage of play:It lowest point in the match was 0.21% when Afghanistan was 99.79%
Forgive me for all the numbers
We see that are 30 overs (partly off screen) we are 6.35% chance of winning. Then in the 31st over we score only 3 runs, well below the required rate, yet our likelihood of winning rises to 7.63%.
But then in the 32nd over, we score 11 runs, Maxwell retains the strike, we've seen off another over of one of the top 10 bowlers in the world, yet our winning probability FALLS to 4.48%. Really?
I suspect there are bugs in this model and it could be that the model is trying to take too many factors into account than data exists for, making it too granular. Duckworth/Lewis works in a similar manner but they smooth that out with an overarching mathematical formula which essentially irons out any of these "kinks".
I also suspect that Cricinfo uses a different forecasting model to that used on TV. The numbers quoted on the telecast regularly differ quite markedly to those being presented by Cricinfo at the same time.
For what it's worth I really like the winning % as part of the coverage, as well as the first innings score predictions, but it's apparent to me they are very much a work in progress.
grumbleguts
User name to post ratio checks out
The actually probability is about 8 balls behind the situation. So it pretty much is fluctuating exactly as you describe the match.To be honest I've grown a bit sceptical of these probabilities throughout the tournament. Consider this passage of play:
View attachment 1851083
Forgive me for all the numbers
We see that are 30 overs (partly off screen) we are 6.35% chance of winning. Then in the 31st over we score only 3 runs, well below the required rate, yet our likelihood of winning rises to 7.63%.
But then in the 32nd over, we score 11 runs, Maxwell retains the strike, we've seen off another over of one of the top 10 bowlers in the world, yet our winning probability FALLS to 4.48%. Really?
I suspect there are bugs in this model and it could be that the model is trying to take too many factors into account than data exists for, making it too granular. Duckworth/Lewis works in a similar manner but they smooth that out with an overarching mathematical formula which essentially irons out any of these "kinks".
I also suspect that Cricinfo uses a different forecasting model to that used on TV. The numbers quoted on the telecast regularly differ quite markedly to those being presented by Cricinfo at the same time.
For what it's worth I really like the winning % as part of the coverage, as well as the first innings score predictions, but it's apparent to me they are very much a work in progress.
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
Hmmmm, you could be right. But if that is the case it's surprising they don't go back and tidy that up after the fact. I mean I just took that screenshot from the Cricinfo app as it appears right now, not how it was in real time during the game.
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
Bevan averaged almost 54... You're possibly looking at his bowling average?
Cleavy
Hall of Famer
Marnus v Marcus: Ponting, Watson want Stoinis dropped for semi-final
Ricky Ponting and Shane Watson want Marcus Stoinis to make way for Glenn Maxwell for Australia’s semi-final against South Africa in Kolkata, arguing that Marnus Labuschagne’s Test match batting skills may be vital in a knockout game.
so drop Stoinis?
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
Yep, and that would be my only change. I don't even think that's controversial; we actually had some decent death bowling yesterday for the first time in YEARS.
Marnus v Marcus: Ponting, Watson want Stoinis dropped for semi-final
Ricky Ponting and Shane Watson want Marcus Stoinis to make way for Glenn Maxwell for Australia’s semi-final against South Africa in Kolkata, arguing that Marnus Labuschagne’s Test match batting skills may be vital in a knockout game.www.theage.com.au
so drop Stoinis?
- Mar 12, 2003
- 3,492
- 5,951
- AFL Club
- Brisbane Lions
- Other Teams
- Brisbane
Looks like I'm in a meeting all day tomorrow
Marnus v Marcus: Ponting, Watson want Stoinis dropped for semi-final
Ricky Ponting and Shane Watson want Marcus Stoinis to make way for Glenn Maxwell for Australia’s semi-final against South Africa in Kolkata, arguing that Marnus Labuschagne’s Test match batting skills may be vital in a knockout game.
so drop Stoinis?
Interesting brettig doesnt bother to list any stoinis career or wc odi stats but just calls him a popular member of the side and an "impact player" whatever that means, this shouldnt even be an argument if we need a batsman who can bowl a bit we go green or we go marnus for batting as stoinis has been awful all wc.
Similar threads
