Pick your nation's 24 best ODI players throughout history

[NZTailender]

I haven't used your simulator but it should take overs per match into account in some way, otherwise you could throw John Parker, Jock Edwards, Warren Stott and Bruce Taylor in and let them run amok. Even with an 75 match cut off Lou Vincent (SR of 20) or Fleming (29, better than 29.2 Bond) would be better than Crowe and Astle.

 

[kiwiviktor81]

I think part of it is that cake players are adding icing elements to their game because of the influence of T20s. Look at Williamson and Guptill - cake players really, but they're both hitting at 85+ recently. Something the simulator showed me was that players like this can really score fast if they're in positions where they don't have to be defensive, and this happens more often when there isn't a collapse up top.

With Guptill and Turner ahead of him instead of Wright, Latham, Greatbatch et al., Andrew Jones hits at almost 70, for example. This is probably how it would have been in real life, as Jones would often be coming in at 70-1 instead of 20-1 and wouldn't have to leave or block so many balls in fear of a batting collapse.

 

[CaptainGrumpy]

What was the best price for NZ v SA? What am I competing against? $1.40? $1.35?

Chasing seems a little harder. Is it 100 games each way?

 

[kiwiviktor81]

I haven't used your simulator but it should take overs per match into account in some way, otherwise you could throw John Parker, Jock Edwards, Warren Stott and Bruce Taylor in and let them run amok. Even with an 75 match cut off Lou Vincent (SR of 20) or Fleming (29, better than 29.2 Bond) would be better than Crowe and Astle.

How it works is that if a player has played more than 75 innings, or bowled more than 75 innings worth of balls (75x60), his stats are left alone.

If it's less than 75 but greater than 50, his stats are nerfed towards mediocrity for someone of their type by 25%.

If it's less than 50 but greater than 25, same applies but 50%. Glenn Turner, for example, who played 34 matches, loses 50% of his extra average over a mediocre player. His average was 47, a mediocre top order player is 30, so Turner averages halfway between 30 and 47 in the simulator, i.e. 38.50.

If it's less than 25, same applies but 75%. In the ATG scenarios this doesn't matter but this is what we did for the 2015 CWC squads.

So Fleming is in the simulator as essentially a non-bowler (EconRate 7.0, BowlSR 100.0). His BowlSr of 29 is therefore nerfed 75% towards mediocrity for a non-bowler, giving an SR of 80 or so.

 

[CaptainGrumpy]

I haven't used your simulator but it should take overs per match into account in some way, otherwise you could throw John Parker, Jock Edwards, Warren Stott and Bruce Taylor in and let them run amok. Even with an 75 match cut off Lou Vincent (SR of 20) or Fleming (29, better than 29.2 Bond) would be better than Crowe and Astle.

heh, I tried using Glen Turner, who bowled one over in his life which was a maiden. The simulator must have something going for it in this regard, because he went for 9 runs an over. I also used Andrew Jones as a bowler. Also not pretty. He was smashed.

Martin Crowe has over a 100 first class wickets and 99 list A wickets at good economy and fair average. He was a very useful bowler. Our poor man's Jacques Kallis.

 

[CaptainGrumpy]

I haven't used your simulator but it should take overs per match into account in some way, otherwise you could throw John Parker, Jock Edwards, Warren Stott and Bruce Taylor in and let them run amok. Even with an 75 match cut off Lou Vincent (SR of 20) or Fleming (29, better than 29.2 Bond) would be better than Crowe and Astle.

I havn't bowled Flemming, but I suspect he will go 9's if not at the very least 7's per over. He will lose far too many games with the ball.

 

[NZTailender]

How it works is that if a player has played more than 75 innings, or bowled more than 75 innings worth of balls (75x60), his stats are left alone.

If it's less than 75 but greater than 50, his stats are nerfed towards mediocrity for someone of their type by 25%.

If it's less than 50 but greater than 25, same applies but 50%. Glenn Turner, for example, who played 34 matches, loses 50% of his extra average over a mediocre player. His average was 47, a mediocre top order player is 30, so Turner averages halfway between 30 and 47 in the simulator, i.e. 38.50.

If it's less than 25, same applies but 75%. In the ATG scenarios this doesn't matter but this is what we did for the 2015 CWC squads.

So Fleming is in the simulator as essentially a non-bowler (EconRate 7.0, BowlSR 100.0). His BowlSr of 29 is therefore nerfed 75% towards mediocrity for a non-bowler, giving an SR of 80 or so.

Cool

 

[kiwiviktor81]

What was the best price for NZ v SA? What am I competing against? $1.40? $1.35?

Chasing seems a little harder. Is it 100 games each way?

The best I ever got was $1.41. It might get a bit harder as I should probably update Rilee Rossouw's stats (he was just about a rookie then and had a very poor start to his career).

 

[CaptainGrumpy]

The best I ever got was $1.41. It might get a bit harder as I should probably update Rilee Rossouw's stats (he was just about a rookie then and had a very poor start to his career).

Was that batting first solely? Or combined legs? I have a chasing team with Crowe batting at 4 doing $1.40 as we speak. Its better when he's down the order.

What was your team? I am interested about the bowlers. I am pleasantly surprised how much difference the bowling order makes even witht eh same group of bowlers. Hadlee is far more an asset at 3 with Bond at 1, than Hadlee at 2. There is no point in him not bowling his ten overs, so where a match would otherwise be lost when bowling second, he keeps the runs down and overs through, with a chance for Bond to mop up the tail for victory in a game that would otherwise be lost. Whoever bowls 5 appears to be protected from I am guessing power play overs. Because they get savaged less than they would bowling anywhere else. Keeps Astle bowling at 5.2, which is managable.

Hadlee is pretty much my Adam Milne and Bond is my Boult.

 

[kiwiviktor81]

Combined legs, i.e. first to 100, then change, then first to 200.

 

[CaptainGrumpy]

Combined legs, i.e. first to 100, then change, then first to 200.

Any thoughts of a test simulator, or does the follow on factor make it difficult?

I could not explain to someone else how much fun your simulator is. They would look at me blankly.

 

[kiwiviktor81]

Yeah Hadlee's much better at 3 because;

if Bond fails, Hadlee can pin down the opposition batmen in the middle overs (usually with Vettori) and prevent the total blowing out of control.

if Bond succeeds, Hadlee can come in when the batsmen are already playing defensively, which doesn't lower his strike rate too much but does lower his economy rate even more, and since it's already 3.3 they're looking at playing Hadlee out for ten overs for around 30.

And Bond's much better at 1 because

his high strike rate means he has a better chance of stopping the openers getting set and milking runs

if all else fails, he can hopefully take out a set batsman when he comes back in over 43 or 44 and stop them from hitting out. You really don't want the opposition to have a top four batsman left in the last 10 overs.

 

[CaptainGrumpy]

Yeah Hadlee's much better at 3 because;

if Bond fails, Hadlee can pin down the opposition batmen in the middle overs (usually with Vettori) and prevent the total blowing out of control.

if Bond succeeds, Hadlee can come in when the batsmen are already playing defensively, which doesn't lower his strike rate too much but does lower his economy rate even more, and since it's already 3.3 they're looking at playing Hadlee out for ten overs for around 30.

And Bond's much better at 1 because

his high strike rate means he has a better chance of stopping the openers getting set and milking runs

if all else fails, he can hopefully take out a set batsman when he comes back in over 43 or 44 and stop them from hitting out. You really don't want the opposition to have a top four batsman left in the last 10 overs.

Hadlee goes for 4.0 for me bowling at 3. I have him with Larsen at 4.7 at 4. Astle goes at 5.1 protected at 5. If he opens for me, he goes at over 5. It just didn't work. Plus he was left with overs when they chased down scores, So I moved him.

Crowe and Bond go for plenty but get wickets. Crowe is currently ont he same wickets as Hadlee. Last night he was getting more.

 

[kiwiviktor81]

Dan wants to do a Test simulator, but I have to figure the AI out first. It's actually really, really hard because you have to account for captain psychology when it comes to team strategy. I.e. a McCullum-led team might bat entirely differently to a Cook-led one if chasing 350 on the final day. They might not. It's really hard to know.

The good thing from a game perspective is that the Test AI would be fiendishly difficult to beat. It would be able to calculate the exact strike rates of your batsmen against its bowlers when setting declarations. It would also be able to calculate exactly the optimum aggression rate when chasing down totals. It would be a demon.

 

[cnerd123]

Wait is this Martin Crowe you are all talking about?

 

[CaptainGrumpy]

Dan wants to do a Test simulator, but I have to figure the AI out first. It's actually really, really hard because you have to account for captain psychology when it comes to team strategy. I.e. a McCullum-led team might bat entirely differently to a Cook-led one if chasing 350 on the final day. They might not. It's really hard to know.

The good thing from a game perspective is that the Test AI would be fiendishly difficult to beat. It would be able to calculate the exact strike rates of your batsmen against its bowlers when setting declarations. It would also be able to calculate exactly the optimum aggression rate when chasing down totals. It would be a demon.

So follow ons and whether to chase down a total or not. Play to win, and ask the user at the start whether to enforce follow ons or not? Declarations will also be a problem. Cannot play udner the assumption of timeless tests these days.

 

[CaptainGrumpy]

Wait is this Martin Crowe you are all talking about?

Yes.

 

[cnerd123]

Zzzz thats what I hate about using sims that only rely on averages. So unrealistic.

 

[kiwiviktor81]

So follow ons and whether to chase down a total or not. Play to win, and ask the user at the start whether to enforce follow ons or not? Declarations will also be a problem. Cannot play udner the assumption of timeless tests these days.

Because there is no bowler fatigue or weather factors in the simulator, the Test version we had of version 2 of this simulator suggested it was always better to enforce the follow on.

Declarations are not that hard with a bit of math. The simulator can look at your lineup and their strike rates and their own bowlers and say "The opposition is likely to score 307 runs off 80 overs, giving us a 47% chance of victory, so if we wait 5 overs we have a 73% chance etc.

Of course, that wouldn't account for what you could do if you played it in Head to Head mode. Still, a well-programmed Test cricket simulator AI would be a nightmare to beat over the course of a series.

 

[CaptainGrumpy]

Zzzz thats what I hate about using sims that only rely on averages. So unrealistic.

What is unrealistic about Crowe bowling? Are you just making assumptions about Crowe being a batsman solely for his career? Are you thinking he was never selected as a all rounder to bowl ten overs in a match?

We don't push your buttons about Bangladeshi players we know nothing about.