International Selections in the game

[6ry4nj]

This is a thread for posting about the similarities and differences between international squads, in the game v. real-life. I have taken my Essex save (ICC09 - same one as in my story) forward to the start of April 2012, and now I'm doing a detailed analysis of the above.

I am currently omitting regens from my discussions, since they are different for everyone, but of course whether you do or not is up to you.

 

[6ry4nj]

Australia

I'm going in alphabetical order, that's why I'm doing Australia first. I may leave Bangladesh for a while, however, because they play so rarely in ICC09.

Anyway, to Australia. Here are their recent real-life squads, with comparisons to my save:

First an explanation of my notation. Where to start?
ITG - in the game
tm - this player is in the team in the game as well as irl
n/a - this player is not in ICC09's database
IFn - International Fail (to be blunt); this player has had n international matches in this format, and has not been good enough
kod - "knocking on door": means they are competitive in that category and will or should be considered for future selection
nwn - "nowhere near": they are never likely to be good enough to select

The numbers in the ITG columns are the players' rankings in the game for that format - by average. Their ranking among international players is used if they have played internationals, and in this case the figure is bolded. So the unbolded figures are domestic rankings.

The averages in blue indicate they relate to bowling, and all-rounders' averages are in purple - batting then bowling.

Thus, for example, Dave Warner has a real-life Test average of 48.3, but in the save he has a batting average of only 35.5 in domestic (having never been capped), the 45th best of Australia's domestic batsmen.

So, here are the game's most recently selected teams, for comparison:

The teams are very different from in real life. In general, the game's selections are more "conservative" than the real-life ones.

In Tests, for example, the game has stuck with Katich and Jaques, rather than experimenting with a Cowan. In this respect it has been more rational than the real-life selectors, as these two (especially Katich) have records which argue strongly for their selection. It has picked up Phil Hughes - you may want to think of him as the Dave Warner of ICC09 (superhuman and can do no wrong). It has also dropped Ponting, although I don't think this is permanent. (Then again, he is the overseas player for lowly Leicestershire...) And it has persisted with Haddin, which even on real-life figures is not that surprising.

It has been somewhat more adventurous with the bowling. It has persisted with Johnson (who is a better bat and bowl in the game than irl), and we are not surprised to see Siddle there - especially with the in-game average close to 25.

I don't understand the selection of Tait. I can only think that the AI knows (or suspects ) the game to have a preference for fast bowlers (ie. over medium-fast/fast-medium). He does have a much better Test record in-game than irl (where he has only played 3 games iirc) - but it's a way off good enough for selection. Nor is his form particularly good.

As for Burt Cockley, in this save he's a domestic star - 4th best career average. Then again, the three ahead of him are McKay, Cameron (a fast bowler btw), and a regen - none of whom are being picked.

So that's the Test comparison done. That's a lot to digest, so I'll post the analysis for the other formats a bit later.

 

[6ry4nj]

Now to ODIs, and again we can make the substitution Warner = Hughes. Again the game has stuck with Haddin, although here the reason appears to be preserving the status quo for its own sake, as Wade?s domestic performance (45.4 batting average in List A!!) makes him look like a better bet.

The game has Rogers, Ponting and Jaques where in real life are Forrest, Bailey and Dussey. Given that Ponting has not had a form slump, I think his selection speaks for itself. Rogers has an impressive ODI average of 37.3 in-game, better than everyone irl other than Ponting, Mussey and Watson, so that?s why he?s there. Forrest and Dussey are not as good in the game as irl, and Jaques has the ODI average in the game that Bailey has in real-life after 5 matches. Bailey?s List A stats ? in the game as in real life ? suggest that his ODI performance might be unsustainable in the medium-term. Anyway, even if his selection makes sense in reality, he?s got only the 31st best domestic average in the game, whereas Jaques has the international runs on the board ? far more than he has irl.

To bowling, and it?s clear why Butterworth and Cockley are selected. More of a mystery is why Johnson, let alone Siddle, gets in ahead of McKay with his average of 21.7. Doherty has no international caps of any description (and that barely-good-enough domestic average), Hilfenhaus has 9 ODIs on his record but all prior to 2009. The Australian spinners are all so weak in the game that it has decided to go without, but there are several pacemen (Bracken, Lee, Tait, in addition to McKay ? to name a few) who ought to have got in ahead of certainly Siddle and even Johnson.

 

[6ry4nj]

Finally, my analysis of T20s. Again we substitute Hughes for Warner ? although Warner is not far from the in-game team in this format. Again we discount the left-field selection of Bailey, who is nowhere in the game. Haddin is the best batting keeper in domestic, Wade is only 3rd after Brummer (that?s what the ?w3? means?). And Haddin is going at 27.67 in T20Is, much better than his real-life 18, while Wade putters along at 20.7 in domestic, nowhere near his real-life 33. So that explains the keeper.

It?s hard to explain why Dussey is not selected (in the game). His 32.3 average is way ahead of Clarke?s and Jaques?. I trust it?s obvious why Pattinson and Doherty are not there. Christian has an in-game domestic average of 12.96, but that?s still only 11th best, and anyway the AI selector hasn?t taken domestic into account. The five bowlers (indeed, the whole team ? except Jaques) are all veterans of at least 4(!) previous T20Is. (No, Christian will never get picked for his batting). Clarke and Jaques are the other two unfathomable batting selections ? both Husseys and Warner ought to be ahead of Clarke. Jaques at least has a domestic average of 30, so maybe that?s the reason in his case.

Apart from Siddle and Watson, the bowling selections are similarly puzzling. Certainly Lee should be ahead of McKay, while just about any domestic player would be better than Butterworth.

----------

As we have seen, some of the AI?s selections are puzzling to say the least. That?s why I?ve come up with these alternative selections here!

I?ve based these purely on international averages, except that I have occasionally picked someone to debut (ie. D Hussey and McKay in Tests, and Wade in ODIs). This means there?s a few selections who have retired or practically retired irl, or who have been mentally ?put out to pasture? by the real-life (and maybe the in-game) selectors. Well, they haven?t retired in the game, so I say pick them!

 

[CG123]

I should be able to do something with my New Zealand save in ICC2005, because by pure coincidence (since I haven't played it in months) the game date is the 1st of April 2012. Since I've been playing as New Zealand and it's on easy, it'll be pretty unrealistic to do an analysis of them. It'll be interesting to see how the teams compare to real life, because the game is now seven years old.

The thing with the AI selections is that they are more based on ratings than performance. I think there is less of a problem with this in the more recent versions, but in something like ICC2005 it's a massive problem. In my Hampshire save Tudor was playing second team for me, and then all of a sudden he got called up to the England team. In my Lancashire save a random regen opener got called up to England in the second season of the game, and he'd only been averaging 30 in both FC and OD cricket in like ten games of each. That is compared to some players who's been averaging well over 30 for a long period of time. In the past I've had youth team players called up to England who haven't played a single domestic game, because even know I know they have a high rating, I like to give the incumbents from last season the first chance, and then when at least one of them struggles I'll bring in the youth team players that I've signed after they've got some form in the second team.

 

[6ry4nj]

England

Here are their recent real-life squads, with comparisons to my save:

I?ve changed the bowling averages to brown as blue was hard to distinguish unless bolded.

And here are the game's most recently selected teams, for comparison:

In Tests, look how the game has jumped at regen batsmen. Still, with Test averages like those so far (and Conran already has 14 Tests), what else could it do? More puzzling is why Bopara and Bell are still regulars. With their game averages under 40, there are three capped players (Cook, Vaughan, McGrath) with Test averages ahead of Bell, as well as multitudes of domestic overachievers led by Patel, Morgan and Cobb (and my - ie. Essex's - star regen Carson, who tops the FC career batting averages with 1715 runs at 59.14). (Patel has been debuted in Tests, unspectacularly, but he is still knocking on the door when in form). The selection of Prior is no surprise ? he has no close competition either in Tests or first-class.

With the bowling, we can see that Bresnan is nowhere near the team. Evidently the AI does not see batting at number eight as a priority, although Swann who bats there does have an average of 25. Of course, before Broad was injured, he was in the team ahead of Bresnan. And where reality has discovered Finn, the game has discovered Woakes instead. While there are so many impressive regen bats, all the top domestic bowlers are non-regens.

The best Test bowler by record is Onions, after 15 matches, followed by Woakes, who has only played 7. The other bowlers who are up there are Kabir Ali, Simon Jones, Sidebottom, Kirtley, Tremlett and Hoggard. Of these six, the computer has only ever picked Sidebottom. If you can understand why those ?blasts from the past? are not being selected, then it is only the omission of Onions that is surprising. The best domestic bowlers are Mark Davies (23.05) and Jonathan Clare (24.08) - and then Woakes. It?s high time Davies or Clare debuted in this form.

So much for the England Test comparison. ODI analysis coming shortly.

 

[6ry4nj]

The thing with the AI selections is that they are more based on ratings than performance.

Yes you're probably right. I just noticed there's only one bat and one bowler difference between the Test and ODI teams (Australia). Nevertheless they are different, and the T20 team slightly different again. Do you know what accounts for the differences? Is it purely a greater emphasis on aggression?

I just had another look through Australia in T20/T20I. The top 8 averages of international bats are all aggressive or very aggressive. The sequence is broken by Mussey in 9th. In domestic however, Mussey is second, and a defensive bat (Katich) is 4th. Surprise surprise, the game never picks Katich internationally in this format despite his top domestic record. It does pick Mussey a fair bit though.

 

[CG123]

Yes you're probably right. I just noticed there's only one bat and one bowler difference between the Test and ODI teams (Australia). Nevertheless they are different, and the T20 team slightly different again. Do you know what accounts for the differences? Is it purely a greater emphasis on aggression?

I'm not sure exactly, but it probably has something to do with that.

Something rather interesting happened today. I was just starting the third season of my Hampshire save, and I'd gone forward to my first match (13th of April) from when it was saved last at the very start of the season (1st of April). It's 2007 world cup year and I have three players picked for England, and they all played in the first game (on around the 10th) with pretty good success. Then as I was selecting my team on the 13th the game encountered a problem (like it does every now and then), so I had to go back to my save on the 1st. I went forward again to the 13th to see that Tremlett (one of my three English players) now wasn't playing in that first game. Now the world cup is well into the super eight stage, and Tremlett still hasn't played at all. I find that very interesting, because essentially in one save he was a starting bowler who originally took 3 for 40 odd in that first game, and then what I can only assume with the exact same rating he hasn't made a single appearance in six games. So I guess based on that there is definitely a random element to their selections.

 

[6ry4nj]

Yes interesting. Did you by any chance take him off bowling practice the second time? That's the only non-random possibility that springs to mind.

 

[6ry4nj]

Now to ODIs (for England), where the game makes just two changes ? Flintoff in for Bell, and Clare in for Anderson. It doesn?t use Trott, even though it has in the past, and his 37 average (from 6 innings) is the 4th best in the format. It has a well-known shortcoming when it comes to selecting non-openers to open, so that explains Pietersen at 3. Callum?s suggestion regarding ratings is the only explanation for picking those two regens in this format ? even in domestic there?s plenty better than them. Cook actually has the third best average in ODIs among openers (after Strauss and Trescothick). If it?s hard-wired not to pick Trescothick (even for home series lol), and only picks openers to open, then it should have had Strauss and Cook. Morgan is the next best ODI opener, but there are plenty of bats ahead of him ? Pietersen, Trott, Bell, Patel, Shah ? even Flintoff. Kieswetter is naturally still a South African Kolpak! The game picks Prior, although Steve Davies is way ahead of him in domestic, with an average of 33.29.

The real-life selection of Patel as 5th bowler looks like a bad idea. In the game he both bats and bowls better. However the AI uses him exceedingly sparingly as a bowler, so he could only be selected in the top six. Regarding Broad and Swann the game agrees with reality. As in Tests, it has discovered Woakes where the current real-life selectors are using Finn (Woakes can?t be far from that role irl). Finn has a good enough domestic average in-game, but nowhere near good enough for England consideration. Anderson, on the other hand, only misses out by a whisker. Jonathan Clare is hugely overrated in ICC09, consequently the game picks him for everything, and frankly he has the international performances to justify it ? primarily with the ball in this format.

Forgot to explain Flintoff, but there's not much explanation needed. He just misses out as a batsman, but makes it as a bowler, assuming the game has decided to retire Wharf and Sidebottom.

And that?s a wrap for England in ODIs. T20Is will be next.

 

[CG123]

Yes interesting. Did you by any chance take him off bowling practice the second time? That's the only non-random possibility that springs to mind.

Na, I did the coaching before I originally saved it on the 1st, because the training doesn't pop up manually until the first lot of games, so you miss out on two weeks of training. He ended up playing 1 out of 9 games.

I had another example of what I was talking about before. This regen pace bowler who was playing for England at the World Cup can't even make my 1st XI. He's only played about 1 out of 5 games since coming back from the World Cup, where he played every match for England, because I pick my teams on form and performance. My bench atm is without doubt the 2nd best team in the competition, behind my 1st XI. I did just lose my first game before though, out of a dozen or so.

Back on topic to this thread, I was having a look before, and on my 2005 save there's really not a lot of point in me doing any analysis. All of the teams only have a couple of players who're currently playing for them in real life. I was looking at South Africa before, and Morne Morkel and Philander are both specialist batsmen in the game... I might do some analysis in a couple of weeks time, just saying how the current real life players are going, but I'm going to be really busy with study over the next few weeks, so I won't both doing anything right atm since it won't be much help for comparative purposes with you and any others who end up doing anything.

 

[deej3005]

It's really interesting to see how the game picks their teams compared to RL, but their selections are obviously flawed (e.g. they continue to pick players even if they've been playing v.badly whereas a human would just choose a replacement)

I'm playing as Australia in my ICC 10 save (the game is now in Jan 2013) and these are the teams that I generally pick (obviously subject to variation depending on the pitch/ form/ opposition)

Test:
Phillip Hughes
Simon Katich
Michael Clarke (c)
Mike Hussey
Marcus North
Shane Watson
Matthew Wade
Mitchell Johnson
Mitchell Starc
Nathan Hauritz
Peter Siddle

ODI:
Phillip Hughes
Mark Cosgrove
Michael Clarke (c)
David Hussey
Mike Hussey
Shane Watson
Matthew Wade
Steven Smith
John Hastings
Mitchell Johnson
Doug Bollinger

T20I:
Phillip Hughes
Mark Cosgrove
David Hussey
Michael Clarke (c)
Cameron White
Shane Watson
Matthew Wade
John Hastings
Mitchell Johnson
Cameron Boyce
Shaun Tait


I just try to pick the best team, without going for realism at all, but my teams clearly vary more than the AI picks would.

If you want, I can post the teams with their ages and averages which would probably be more interesting.

 

[6ry4nj]

If you want, I can post the teams with their ages and averages which would probably be more interesting.

Yes if you'd like to share, I'd like to know:
- the career averages of the players you pick
- the top ten international bats by career average in each format
- the top ten international bowlers (and top spinner if not in top ten) by career average in each format
- the top two intl keepers by career batting average in each format
- the top five domestic bats (by career average) in each format
- the top five domestic bowlers (and top spinner if not in top five) by career average in each format
- the top three domestic keepers by career batting average in each format

...and their career averages.

If you're too busy, any subset of that will be appreciated. The easy way - for batsman and bowlers anyway - would be to attach screenshots.

I'm particularly interested that you pick a spinner in each form. In my ICC09 save, Australia hasn't had a spinner perform in any format for three years. Do Hauritz/Smith/Boyce perform for you?

Also regarding Johnson and Starc in Tests. I suppose they must go good or you wouldn't pick them. Johnson certainly performs unrealistically well in ICC09 (and only in Tests, whereas irl I believe he's much better in limited overs).

What's the deal with Marcus North? He's a has-been in ICC09...

Yeah - lots of questions! Anyway, thanks for joining the discussion!

----------

Back on topic to this thread, I was having a look before, and on my 2005 save there's really not a lot of point in me doing any analysis. All of the teams only have a couple of players who're currently playing for them in real life.

It's still worth doing the comparison, even if there are no players in common to the real-life teams - assuming they're not all regens and incorrectly databased players that is... You can still say why the real-life players aren't getting picked (assuming there's a valid reason ), and I can look up the non-regen players who are playing, and marvel at how wrong the game got its predictions.

Of course, study comes first. Good luck with that!

 

[CG123]

It's still worth doing the comparison, even if there are no players in common to the real-life teams - assuming they're not all regens and incorrectly databased players that is... You can still say why the real-life players aren't getting picked (assuming there's a valid reason ), and I can look up the non-regen players who are playing, and marvel at how wrong the game got its predictions.

Yep, I'll definitely try to do something. In most cases the teams have a handful of players who are also there in real life, a handful of regens, and a handful of real players who are no where near that good in real life. Because it's like seven years old, most players who're under-25 in real life aren't there, which is where the main difference in the teams lie.

Of course, study comes first. Good luck with that!

 

[deej3005]

Here is a bit more detail on my Team first-the brackets are age, then Test batting avg., then Test bowling avg if they have one:

Test:
Phillip Hughes (24) (52.40)
Simon Katich (37) (49.14) (37.43)
Michael Clarke (31) (54.31) (44.73)
Mike Hussey (37) (49.44) (63.50)
Marcus North (33) (42.82) (57.11)
Shane Watson (31) (45.28) (31.77)
Matthew Wade (25) (39.62)
Mitchell Johnson (31) (26.27) (28.40)
Mitchell Starc (23) (23.00) (26.85)
Nathan Hauritz (31) (18.51) (32.15)
Peter Siddle (28) (14.42) (27.79)

The top 3 of Hughes, Katich and Clarke are easily my best bats. As you can see we have a bit of an age problem, batting wise since I pick my teams only with the current series in mind and not the more distant future.

When Simon Katich retires, he will probably be replaced by Mark Cosgrove (who is 28 and has 4 Tests at 38.57). It's not so easy to see a replacement for Mike Hussey though...Steven Smith started amazingly with a bat avg. of 131.50 after 3 Tests but he then fell away shockingly, so that now he avgs 29.70 with the bat from 13 Tests, and 37.41 with the ball. Matthew Wade also started really well and has since fallen away, but to a lesser extent. Marcus North is definitely my weakest bat, but has by no means been a flop for me.

We've got no age problems bowling wise, the youngsters are doing superbly. Starc has made a great start, Johnson does far better than irl, Siddle pretty much the same as rl, and Nathan Hauritz has surprisingly been excellent. There are a number of bowlers knocking on the door as well, especially Peter George (26 and avgs 29.51 from 15 Tests) but also Doug Bollinger, Andrew MacDonald (the all-rounder) and Burt Cockley/ Luke Feldman. No other spinners though, only Cullen Bailey did half-decently for a bit...

ODI:
Phillip Hughes (24) (46.18)
Mark Cosgrove (28) (41.94) (69.00)
Michael Clarke (31) (44.27) (36.34)
David Hussey (35) (33.40) (111.00)
Mike Hussey (37) (50.42) (102.50)
Shane Watson (31) (42.49) (26.30)
Matthew Wade (25) (20.80)
Steven Smith (23) (26.94) (29.45)
John Hastings (27) (17.20) (21.03)
Mitchell Johnson (31) (18.28) (27.13)
Doug Bollinger (31) (0.33) (19.36)

Again batting wise my openers (and no.3) are my best batsmen. Michael Clarke is a rock in both formats, Phil Hughes is obviously amazing (though not quite as amazing as he used to be in ICC 09). The middle-order is a bit more iffy, David Hussey has batted well, he just started with a very low average (about 31.00) but he can't bowl at all and Cameron White is right on the borderline of the team (he avgs 34.79 with the bat and just 24.19 with the ball but is really inconsistent).

Matthew Wade hasn't quite taken to this format like he has in Tests, to the point where Haddin might be brought back in soon. Steven Smith plays but as a bowling all-rounder, although his bowling is not that good (economy 5.28) and John Hastings is my star with the ball, Mitchell Johnson can be destructive but is inconsistent (like Shaun Tait). Again, there are many more options with the ball, Shaun Tait avgs only 23.57 but is unreliable, Cam Boyce avgs 19.73 and is a spinner but hasn't played for more than a year since he disappeared from the game (I think his contract wasn't renewed by Victoria). Then there's Mitchell Starc, Peter George, Ryan Harris, Burt Cockley and possibly Nathan Hauritz.

T20I:
Phillip Hughes (24) (34.45)
Mark Cosgrove (28) (25.60)
David Hussey (35) (28.29) (36.29)
Michael Clarke (31) (38.65) (30.50)
Cameron White (29) (26.48) (34.50)
Shane Watson (31) (37.32) (21.84)
Matthew Wade (25) (38.00)
John Hastings (27) (30.50) (23.23)
Mitchell Johnson (31) (12.67) (16.25)
Cameron Boyce (23) (--.--) (40.50)
Shaun Tait (29) (7.00) (16.79)

I tend not to take T20I's too seriously since they're never a full series, but Hughes and Clarke are again beasts (they're the only two 'undroppables'). Strangely enough, Wade is really good at T20I despite being so far poor in ODIs. T20Is clearly seem to favour more aggressive batsmen though, which is why White and Watson do well (while Mussey failed), though White for some reason can't bowl here.
Mitchell Johnson is my best bowler, with Shaun Tait not far behind which makes me think T20Is favour the quicks, though Bollinger is also v.good. Hastings' speciality is keeping the runs down rather than taking wickets (as in ODIs) and Boyce is being 'eased' back into international cricket through T20Is after his year-break but only has 2 wickets and has never batted.
Sometimes I like to experiment in T20Is but few of the newcomers introduced have done well, except Cockley, Harris, Feldman and (maybe) Voges.


Whew, that's that done. I'll get on to the other stuff tomorrow.