Quick question regarding skill stats and effects.

[enigma]

Is there an updated roster from someone intelligent like yourself? I don't think I can handle bowling Gillespie for Vettori anymore lol.

No, I haven't made any rosters but I think Chewie had it on planetcricket so you may want to have a look at that & if you like how he's rated players then go for it.

I was thinking about making one but the thing is, if you change ratings of some players & not everyone in the game then that could lead to some statistical imbalances & inaccuracies as I've said before, like your own batsmen/bowlers performing too well which then becomes unrealistic; for the same reason, I remember debating Chewie a while back as to why, under the circumstances (with spinners overperforming,etc), Vettori should NOT get his deserved rating of ~650, under an updated roster.

So as far as I'm concerned (although others may think otherwise), either all players in the game should be edited/adjusted or I'd rather not bother with it, otherwise we're back at square one ie some players overperforming (be it your own) & some underperforming; & the thing is even editing everyone mayn't get you the best results because playing conditions aren't realistic enough. May be I'm way too persnickety about this but being a stats-maniac I can't help it.

For me, the only way to fix this issue is getting the playing conditions to resemble real-life as closely as possible. If you observe closely, all the players' real-life-statistics are massively affected by playing conditions, so if the playing conditions are gotten right (of course, there'll've to some leeway)then all players can be given a rating close to what they deserve & yet their stats will be real-life-like. Right now, as I've said in the other thread I've started, playing conditions are neither sufficient nor are they having the proper impact on the game & this short-coming then has to be compensated for by underrating/overrating players; for example, since spinners are overperforming in the game, Vettori must be given a poor rating or else he'll average 25-something if given he's given the rating he deserves (similar thing with Malinga's rating of ~800), the situation is exactly the opposite for example for Mark Gillespie, Tuffey, etc

Us editing a few players isn't going to make the game-play drastically better, the root of this issue is in the playing conditions, if they're gotten right then most things will automatically fall into place.

Seconded. I have no intention of cheating I'd just like everything to be a bit more realistic. I hate playing against SA a couple of years into the game cos by then their team is full of the 'awesome' players like Pelser etc. who only average 30-odd in the top flight FC in SA.

I think ICC-developers should've separated the performances of SA players between professional & amateur competitions & should've based the ratings only on their professional competitions. It'd be a very tedious job to go to cricinfo & record the professional-competitions' records separately for each season for each SA-player but certainly not an impossible one but I suppose they didn't pay close attention there because South-Africa-Domestic isn't playable in the game but as I've said, ratings are all relative & having some overrated/underrated players affects the statistical balance of the whole game.

If you've read any posts here about, a few seasons down the line, all the best batsmen averaging in the 30s & bowlers averating in low 20s or all the best batsmen averaging in the 50s & bowlers close to 30 then that's another instance of imbalaced ratings, only in this case, imbalaced ratings of the regen-players.

[ab5ides1]

I think ICC-developers should've separated the performances of SA players between professional & amateur competitions & should've based the ratings only on their professional competitions. It'd be a very tedious job to go to cricinfo & record the professional-competitions' records separately for each season for each SA-player but certainly not an impossible one but I suppose they didn't pay close attention there because South-Africa-Domestic isn't playable in the game but as I've said, ratings are all relative & having some overrated/underrated players affects the statistical balance of the whole game.

If you've read any posts here about, a few seasons down the line, all the best batsmen averaging in the 30s & bowlers averating in low 20s or all the best batsmen averaging in the 50s & bowlers close to 30 then that's another instance of imbalaced ratings, only in this case, imbalaced ratings of the regen-players.

Another thing I've noticed is that sometimes players average 20-35 in test cricket and all their big scores are made against me, as in they 'up their game' when playing against me, but only against me. And yeah, by 2020 my batsmen all have far-and-away the best batting averages, but my bowlers have normal bowling averages that aren't uber-good like sub-20. So I end up with easily the most realistic team.

[enigma]

Another thing I've noticed is that sometimes players average 20-35 in test cricket and all their big scores are made against me, as in they 'up their game' when playing against me, but only against me.

Do you note down who scores against you, etc because unless that's done, it's kind of hard to say if it is actually a bug. May be it's just coincidence, I'm sure there are enough of those 20-35-average-guys who score very little against you & therefore go unnoticed; again, it could be a bug but I can't say I've personally seen enough of this to be able to assert that it is.

And yeah, by 2020 my batsmen all have far-and-away the best batting averages, but my bowlers have normal bowling averages that aren't uber-good like sub-20. So I end up with easily the most realistic team.

I meant best batsmen/bowlers in the whole game (not just the ones on your team) & by your description, it seems your team falls into the first category of better batters & average bowlers. If your batsmen have the best averages amongst the game then you probably have batsmen with unrealistic batting-ratings, I wouldn't be surprised if some of those ask for a few hundred-thousands at the end of their contracts. You know, I'd once seen a batter who'd asked for 5.2 MILLION (yeah, that's right ), averaged 1000+ in all forms of the game, his batting-rating was ~56000 with a potential of improving upto the maximum of ~65000 so in his early 20s he was 35 times as good as Ponting

Regens like these make the game unrealistic at times, I think batting/bowling-ratings should be capped around about where the original real-life-players' ratings lie. These super-regens are put into the game to create an illusion that somehow we're getting better at managing our side but that's usually not true, it's just that we often have unrealistically good players, that's why I usually like playing only the first few seasons. Not to mention, if regens are capped at realistic level then that could help somewhat with online-cheating as well as the server can just throw out any team having players with ratings above the capped levels.

[6ry4nj]

I think ICC-developers should've separated the performances of SA players between professional & amateur competitions & should've based the ratings only on their professional competitions. It'd be a very tedious job to go to cricinfo & record the professional-competitions' records separately for each season for each SA-player but certainly not an impossible one but I suppose they didn't pay close attention there because South-Africa-Domestic isn't playable in the game but as I've said, ratings are all relative & having some overrated/underrated players affects the statistical balance of the whole game.

Just had a quick look at Pelser. He plays FC for one team (North West) in one comp (now called CSA Provincial 3-Day Comp - only looked at FC). Are there professional and amateur teams in this comp? Maybe he really is good! (FC batting avg of 62 off 33 completed innings with aggregate just over 2000) Better (average) than Ponting or Tendulkar - although obviously he has no Tests included in his. Maybe we need first-class averages that exclude Tests to compare apples with apples (this way Tendulkar's average is 65.39!).

Cricket Archive has records by player split by competition, year, even opposition. Check it out.

[Sureshot]

Right, might be a bit long this...

Pelser and other players who have played in the CSA provincial system, this gets classed as first class cricket, but is more like the level of our Minor Counties or perhaps nearer Associate cricket. This inflates their stats and we'll be introducing some adjusters in to 2011 to negate this.

Are there other countries where there are similar systems and it has a similar affect on players in the game?

If you notify us, we will improve how it all works and make it more realistic. I've been a bit under the impression that SA domestic is a bit unusual in this regard, but maybe I'm wrong.

Ratings are very difficult to get right, whilst I will not disagree on the fact the Pelser and co are obviously wrong (and we will rectify this), a lot of the other ratings are subjective.

We'll introduce new systems this year, that will improve player ratings, I'm not going to claim we will get every one right, because a) in some cases it's subjective and b) with the sheer quantity of players, mistakes will happen, we want to reduce this, but you'll never get rid of it.

With regards to how the ratings are made or how they work in the game, I'm not going to comment on it, I understand this may be frustrating, but there are various reasons for it. Which are also why we've been a bit quiet on it before, but we probably could have made this statement before.

Taking that statement in to perspective though, we will strive to improve the ratings and make the game more realistic, we will learn from our mistakes and look to make the game better for every one involved.

We will look to make improvements in other areas over the coming versions as well to make it easier to select players, a variety of "selection tools" is our aim, but this is something that will take time to develop and is going to be built up over the coming years. Not only due to time constraints, but we can improve it with your guidance. I've input quite a lot of my own ideas for these tools, it's absolutely the right direction, viewable player ratings isn't something we will do, but we also understand that we need to improve how you can tell which player is better than others.

The long term goal of this, is that in the future (because it will take time for us to develop, analyse and improve) you will have a wide expanse of stats to help you analyse how your players are performing. Again, this is a long-term goal and not short-term.

I'm a statto (I regularly look at England records before our matches to see if someone is about to break a record), I think most players of ICC are as well all of us developing it, there's room for improvement, and we will get there.

I'll try and answer any questions on the above if you've got them.

I think it's good that the forum opens up this and we really do value all your input, I just hope you understand that there will be things we can't comment on and that others we can. The things we can (and we can talk about what players that need improving/nerfing) we will look to reply in the best way we can.

Edit:

Forgot to mention our position on an official editor. It's not something we are planning for 2011, in any form.

[Xeno]

Thanks for the reply Sureshot! Definitely a helpful discussion on the progress planned for the future, exciting stuff!

I guess the only downer for me in the entire game is the ratings, which it seems like I can easily fix for my personal game. So no biggy for me. Thanks for the discussion

[6ry4nj]

The first and very important question I have is:
Are you seeking to make it impossible for third-party editors/viewers to work with the game?

I understand that no editor has been found that works with the 'final' release of ICC10. But the community only found the ratings errors because some of them had editors/viewers that worked (on earlier releases). If the 2011 version has no compatible editors/viewers, we will never again find a ratings error.

The second question is: what is stopping you from fixing the 'howlers' when they are pointed out?

How can fixing Pelser take a year? It is a question related to the first thusly. If you slow down the rate at which errors are found (by excluding the community from the process), and don't speed up the rate at which they are fixed, then the current cycle of disappointment could well continue for a decade or more.

Even without an editor, after a few months the players will report that someone is insanely good in the game (probably two or three as they did Pelser, Hughes, Flowers...etc? in this one). Following June, three more fixes, and by August or so, three more howlers discovered.

The third question is: did you take my point about taking out Tests when comparing first-class averages? When someone has played the majority of the first-class matches as Tests, their averages will tend to be worse than they would be if they had only played domestic. If you compare everyone (as a first step) on the basis of their domestic performance, you have something closer to an apples-with-apples (as opposed to chalk-and-cheese ) comparison.

[Chewie]

There is no editor for the final one (that I know of) because we (at PlanetCricket) told people not to make one. If they want to edit the game they can do so on an unpatched version of the game which is not able to be played online - I felt this was a valid compromise.

One option is having a special option when starting a new game which determines whether a built-in editor can be used, and this game cannot be used online.

[enigma]

Just had a quick look at Pelser. He plays FC for one team (North West) in one comp (now called CSA Provincial 3-Day Comp - only looked at FC). Are there professional and amateur teams in this comp? Maybe he really is good! (FC batting avg of 62 off 33 completed innings with aggregate just over 2000)

I averaged 100+ when playing with my mates, may be I was really good too!
Jokes aside, the thing is that almost all of the SA-players that represent the national side have played in the SuperSport Series which is much is more competitive & professional & if Pelser, by the age of 26 hasn't played a single FC in that Competition then he's likely not that good.

Secondly, statistics produced in 3-day competitions tend to be different than those produced in 4-day competitions because 3-day-games usually have a tendency of ending in draws so some people end up playing for stats hence inflating their stats while some just take them as "for-fun" matches & play casually hence their stats may be a little worse off for that. Not to mention, when teams do try to force results in 3-day-games then that leads to further distortion of statistics. So in short, even though both are classified as "First-Class", 3-day & 4-day competitions end up with different statistical balances.

Moreover, 3-day-CSA-Provincial has 14 teams which means more average players can get "First-Class" matches in while SuperSport Series has only 6 teams which means the talent-pool is squeezed & only the better ones get to play that competition & hence higher quality of cricket is produced.

Better (average) than Ponting or Tendulkar - although obviously he has no Tests included in his. Maybe we need first-class averages that exclude Tests to compare apples with apples (this way Tendulkar's average is 65.39!).

At first look, this seems very logical & proper BUT the thing is, once a player becomes an established International, they play very little domestic-FC (as they're saved for International matches), may be a couple of games here & there when they're out of practice or form which means that Ponting or Tendulkar would've'd even better domestic-FC averages had they not played any International-cricket, not to mention, with guys like these, they'll've played most of their domestic-FC when they were younger, less experienced, less mature & hence these stats won't reflect their current true worth which makes such a criteria even more flawed.

Moreover, a perfect example of why the proposed criteria is flawed is Dale Steyn, his Test-average is 23 while his domestic-FC-average is 25 so are you saying that anyone who has a better average in domestic-FC should get a better rating than Steyn! I guess not. And if such a criteria is applied then that'll mean that ICC have been aboslutely right in rating the likes of Masekela, N Miller, C McKay way better than Steyn

In short, I think it was a good proposition but there'll be issues with that as well. I believe that established Internationals should be rated on their International performaces.

EDIT : I was just going through some statistics & just realised that Sangakkara's average drops down to a meagre ~37, if we exclude his Test-stats from his FC-stats so there you go.

[enigma]

Are there other countries where there are similar systems and it has a similar affect on players in the game?

If you notify us, we will improve how it all works and make it more realistic. I've been a bit under the impression that SA domestic is a bit unusual in this regard, but maybe I'm wrong.

To be honest, countries that don't have a single, unified FC-Competition (Eng, Pak, SL, Ind & of course, SA) all suffer to various degrees with the issue of some of its better players ending up with better domestic-FC-stats than they deserve.

How about Eng itself, are Division-Two FC-stats worth as much as Division-One? May be I'm being persnickety but there definitely is a difference between the FC-stats of better players in Division-One & Division-Two; with two players of equal quality, one who's played a lot of Division-Two will generally have better FC-stats, thus basing the ratings on overall FC-stats will always lead to overrating/underrating of some players, if the differences in quality of competitions aren't taken into account.

It's obviously impossible to account for each & every factor that may inflate/deflate FC-stats but two obvious things that do so are :
1) 3-Day-Games : As I've mentioned to 6ry4nj in the post above, 3-Day-Games can't be put on the same level (for many of the reasons why we put long-form & short-form stats separate), 4-Day-Games should be given higher consideration than 3-Day-Games when it comes to determining ratings
2) More teams : The more teams there are, the more average-players get into the competition which then leads to good-players' stats getting inflated as the quality of cricket becomes less competitive. One of the reasons Aus have the highest quality of domestic-cricket, it is because talent-pool is squeezed into only 6 teams with a single unified competition which produces for more competitive cricket.

Considering the above two points, CSA-Provincial obviously gets a tick on both counts & hence the inflated stats.

Srilanka's Premier League Tournament has 20+ teams & it's a 3-Day-Competition (it's basically club-cricket) so it gets a tick on both counts as well, not to mention it's a tiered system so another tick while the other FC competition in the country has only 5 provincial teams & is 4-Day-competition which means better quality.

Pakistan is almost identical to SL except both competitions are 4-Day but 20+ teams in one of the competitions, dilutes the quality & inflates FC-stats of better players while the other seems better with 5 teams only.

For India, with its massive populace (& interest in cricket), one might say 25+ teams isn't "way too many" but I'd say enough to dilute the quality of cricket somewhat, so Ranji is OK but not so great (obviously, better in SuperLeague than in PlateLeague). The other FC-Competitions Duleep & Irani (one-off 5-Day-Game) are pretty short but quality is better as the talent-pool is squeezed into 5 & 2 teams respectively.

Having said all of this, as I've said numerous times before, the biggest factor that has an impact on statistics is playing conditions so until they're accounted for (of course, it'd be near impossible to make a "perfect" system) & the playing conditions within the game are brought on parity with real life to the extent possible, it'd be almost impossible to make completely realistic game.

[Sureshot]

Thanks for the reply Sureshot! Definitely a helpful discussion on the progress planned for the future, exciting stuff!

I guess the only downer for me in the entire game is the ratings, which it seems like I can easily fix for my personal game. So no biggy for me. Thanks for the discussion

Player ratings is much higher on my priorities this year, it will be improved. In accordance with Enigma's very enlightened post above, I have a better understanding of the other domestic leagues as well. So I'll discuss this with Chris, we've also agreed to look at spin and it's impact.

When I was researching South Africa I noticed that Wayne Parnell hadn't played for any of the franchises, so he's an exception. Does anyone know of others?

Another country that would be good to know more about domestically would be the Bangladesh system. A lot of their spinners have good FC stats, over-inflated for sure.

If we can build up a profile of domestic leagues that have 3-day cricket and this lower standard, then we will work with it in the code. With these Sri Lankan sides, I'll do a bit of research in to what teams are the provincial ones that play the higher standard. Like in South Africa, we can code it so those that play for the franchises, their stats mean more.

Enigma, thank you so much for your feedback above.

Improving the dynamics of the system, is definitely something that will take time, but I'm aiming for a good improvement this year.

Even without an editor, after a few months the players will report that someone is insanely good in the game (probably two or three as they did Pelser, Hughes, Flowers...etc? in this one). Following June, three more fixes, and by August or so, three more howlers discovered.

Can't answer the first part, but, I'll answer this.

Been looking at my code and I did substantially nerf Pelser, maybe it wasn't enough (I think I realise now his standard of cricket better than I did), I used to be under the impression he was one of SA's better young batting prospects, but it's clearer now that he isn't.

Hughes, again, we nerfed, we'll review it again. Maybe we didn't nerf him enough, but it was quite substantial

Calvin Flowers? Not heard this before, I see he's a Saffer in the DB.

As said, we've got new systems that will come in to place for 2011 ratings creation, they are taking in to consideration everything said. We'll take on board all your feedback, and we can analyse it (both you and us) once we've released, the objective is obviously to have a greater set of accuracy and get rid of the howlers. I think subjective ratings will always exist, but it's a priority to remove the howlers.

I've got a better understanding of the standards of the global domestic cricket scene now, which is very helpful.

[Blade007]

When I was researching South Africa I noticed that Wayne Parnell hadn't played for any of the franchises, so he's an exception. Does anyone know of others?

Parnell Plays for the Warriors

[enigma]

The first and very important question I have is:
Are you seeking to make it impossible for third-party editors/viewers to work with the game?

I hope NOT because....

If the 2011 version has no compatible editors/viewers, we will never again find a ratings error.

If you slow down the rate at which errors are found (by excluding the community from the process), and don't speed up the rate at which they are fixed, then the current cycle of disappointment could well continue for a decade or more.

....Exactly. If the ratings are made unviewable then such mistakes will never be pointed out & that will be a loss forthe game & its developers as people won't keep on buying the game if their real-life heroes don't perform as well they should while some unknowns end up being the best performers in the game; not to mention there're other aspects of the game which need improving & it'd make it very difficult to point it out if "entrails" of the game can't be viewed. The only reason this massive flaw of the game was pointed out is because at least some people could see the "entrails". If ratings are made completely unviewable then that'll only retard the development of the game & in turn, lose customers. There're better ways to deal with online-cheating than making ratings unviewable.

The second question is: what is stopping you from fixing the 'howlers' when they are pointed out? How can fixing Pelser take a year? It is a question related to the first thusly.

Well, they did reduce Pelser's rating somewhat after the excel-sheet was released. I believe he was 1745 after the initial release of ICC2010, after a couple of patches, he was 1788 (this is when we'd compiled the excel-sheet) but after the excel sheet was put up for everyone to see, in the next patch a few days later, they'd reduced him to 1426, still pretty high I'd say but most of the others weren't reduced, in fact, if I remember correctly, Nikita Miller (WI) who was previously 561, was BETTERED to 507 (bowling God )

When I was researching South Africa I noticed that Wayne Parnell hadn't played for any of the franchises?

When was this? He's been playing for Warriors since as early as 2008.

Another country that would be good to know more about domestically would be the Bangladesh system. A lot of their spinners have good FC stats, over-inflated for sure.

Well, Ban may seem like a conundrum at first as they play 4-Day-Cricket AND have only 6 teams but then enters the factor of pitches. They've very slow dead pitches which neither encourage good batsmanship nor good bowling, such pitches only encourage "slow bowling" (whether spin or medium/cutters-type); ball doesn't come on to the bat which means batting is difficult & that's why if you go to their "Player Records" page in the game & sort their players "By Bat" then you'll see poor batting averages & hence their slow-bowlers' averages are better than they should be; same will be the case to various degrees with WI, SL, it's the result of slower pitches; at least SL has a couple of decent pitches that help produce some decent batsmen but if you sort SL-players' list "By Bowl" you'll find hordes of spinners with really good averages due to their slow low pitches & uncompetitive 3-Day-Competition; Ind & Pak's pitches aren't as slow as Ban, SL, WI & that's why they produce decent batsmen & spinners (quality spinners always prefer SOME bounce & carry) & you'll also find that vast majority of Ind/Pak's better-averaged (25 & below) bowlers are actually seamers, whether medium-swing or fast; WI-cricket's woes lie in slowing down of their pitches too, you'll see very low batting averages if you sort them "By Bat", WI in their pomp had faster, bouncy pitches more or less like Aus & SA which encourage good batsmanship & mostly good pace-bowling, & pitches are the reason why Aus & SA have had domestic-seasons of the highest standards for a long time.

If we can build up a profile of domestic leagues that have 3-day cricket and this lower standard, then we will work with it in the code.

Currently, I think only SL & SA's 3-Day-domestic is being classified as FC along with their 4-Day-domestic but as I've said before it's not just about 3-Day; just because a competition is 3-Day doesn't in itself make it poor, my point was that stats produced in 3-Day-Games are different from those produced in 4-Day-Games, the fact that SA & SL's 3-Day-domestic are poor compared to their 4-Day-domestic largely comes down to their 3-Day having many more teams & hence good players being able to beef up their stats on the backs of average-players that are allowed in due to more teams competing. And mentioned before, players' stats can also "deflate" (worse off) due to playing in 3-Day because as some mayn't take them seriously or if some batsmen have to throw their wickets in search of quick runs in order to force a result in the game.

Further, two different 4-Day competitions within the same country may be of varying standards so it may be necessary to take that into account as well rather than treating them as equal, not to mention varying differences in standards of 4-Day cricket between different countries.

May be it's worth mentioning here that until 1990 or so, most countries' domestic-competitions were mostly 3-Day (except Aus & WI who'd 4-Day-domestic for a long time) but once the benefits of 4-Day were recognised, some teams (Pak,Eng,Ind,NZ) turned their 3-Day domestic-competitions into 4-Day competitions while some (SL,SA) left their 3-Day competitions in place while separate 4-Day competitions were formed.

Enigma, thank you so much for your feedback above.

No worries. Always happy to help.

Calvin Flowers? Not heard this before, I see he's a Saffer in the DB.

I think I've heard a fair bit about him running through county sides (within the game, of course) on this forum as well as on planetcricket & why not, he also had a bowling-rating of around about 650 if I'm not mistaken & the fact that he's a spinner would make him even more lethal.

Improving the dynamics of the system, is definitely something that will take time, but I'm aiming for a good improvement this year.

As said, we've got new systems that will come in to place for 2011 ratings creation, they are taking in to consideration everything said.

Avidly looking forward to the changes.

[Chewie]

Re Calvin Flowers, if you look at my Worcestershire ICC story (With CG123 and Hedger_14), Hedger had him and that was the main reason why he did so much better than us.

Re unviewable ratings, that'd basically be impossible. There is one basic way to view ratings that doesn't require an actual editor as such, and I doubt they'd change the way the game is run so much that it wouldn't be possible to view the ratings through this method. As far as I know, I'm one of the only people who still know how to view them this way though. And I have to half reteach myself each time I want to do so

[Themer]

As far as I know, I'm one of the only people who still know how to view them this way though. And I have to half reteach myself each time I want to do so

Quite alot of people know. Anyone involved in the making of the editor had previous experience. When Fe was doing it there was still quite a few people whom knew and anyone who's done it for any other version knows how albeit they did move around after a few versions.