cricrate: new cricket ratings website

[fredfertang]

Have a read of this viriya

 

[Red]

Did you see my post above, the reason Imran is rated lower than them is because his career was mostly solely as a bowler early on and then more as a batsmen later. During later years there were times when he didn't bowl and just batted. I agree that he ended up with a great all-round record, but he didn't have consistent all-round performances because of this. If he gets a 100 and doesn't bowl in a match, his all-round rating for that match is 0, because it wasn't an all-round performance. Likewise if he barely bats and just bowls, same thing happens.

This rating system is bottom-up, so the overall averages do not figure in the rating system - it takes into account each match on its own.
The reason why Shakib and Cairns are figured higher is basically for the opposite reason - for most of their careers they were genuine all-rounders.

This method of multiplying batting and bowling ratings for a match works very well in general - since all the top all-rounders (except for Imran) figure at the top. That Imran is penalized was not planned but I think the reasoning is not entirely unfair.

This is not to say that I think Imran is an inferior all-rounder - just that his career was not full of all-round performances but separate bowling and batting brilliant periods. I am aware though that there was a period where he was dominant with both, but that did not nullify the effect of his career being disjointed.

This is still a good argument to add a factor like batting average/bowling average in addition to the average all-round performance though - so a player like Imran (who was obviously a great all-rounder), is not penalized excessively for starring in one discipline a match frequently.

Ok. That makes more sense then.

I do think that you need to make that very clear on the page. Because people are going to look at that list and go "wtf?"

 

[viriya]

Ok. That makes more sense then.

I do think that you need to make that very clear on the page. Because people are going to look at that list and go "wtf?"

That's why there's a Methodology link right on the top of the page
cricrate | Test All-Round Ratings

 

[viriya]

Mike Procter the 89th greatest all-rounder?

Darren Sammy above Trevor Bailey?

I can see where LT is coming from tbh

The reason Mike Proctor is not rated higher is obvious - he only played 7 matches. Even though we know he was a great all-rounder he suffered from SA being excluded from international cricket and unfortunately there is no way to resolve this statistically (same goes to Barry Richards).

Sammy vs Trevor is not obvious to me either. Just from a quick overview I think Trevor's career is affected by batting during a weak bowling period (check Hobbs/Hutton discussion above), and also his great bowling performances not coinciding with adequate batting performances (and hence not all-round performances - similar to the reason Imran isn't rated higher).

 

[viriya]

More in-depth on Bailey's career:
cricrate | Trevor Bailey

In his three best bowling performances according to the rankings:
7/34 vs WI
7/44 vs WI
5/20 vs SA

he didn't really feature with the bat - his highest score of those 3 matches was 23. So all three of these matches are not considered great all-round performances even though they were great bowling performances.
Similarly, in his 3 best batting performances:
138* vs NZ
88 vs AUS
82 vs WI

his best bowling figures were 3/140. So looks like he went through the "Imran effect" as well.
You can check the match scorecards from the innings lists.

 

[viriya]

Have a read of this viriya

interesting - i'll take a look.

 

[fredfertang]

More in-depth on Bailey's career:
cricrate | Trevor Bailey

In his three best bowling performances according to the rankings:
7/34 vs WI
7/44 vs WI
5/20 vs SA

he didn't really feature with the bat - his highest score of those 3 matches was 23. So all three of these matches are not considered great all-round performances even though they were great bowling performances.
Similarly, in his 3 best batting performances:
138* vs NZ
88 vs AUS
82 vs WI

his best bowling figures were 3/140. So looks like he went through the "Imran effect" as well.
You can check the match scorecards from the innings lists.

I see what you mean now - but not a atatistic that I can see is of any use

 

[viriya]

I see what you mean now - but not a atatistic that I can see is of any use

I see your point though - I think I'm probably gonna add a separate factor when calculating all-round career rankings based on the battingAvg/bowlingAvg ratio which would tweak the ratings a bit.

 

[Red]

I think the combined performance in particular matches is an interesting stat, but not a very comprehensive one.

I'd be very interested in who's combined runs and wickets most successfully over a series or over a 12 month period.

 

[viriya]

I think the combined performance in particular matches is an interesting stat, but not a very comprehensive one.

I'd be very interested in who's combined runs and wickets most successfully over a series or over a 12 month period.

most likely it was imran khan - he had a period (something like 10-15 tests) where he averaged 50+ with the bat and <25 with the ball.

 

[Lillian Thomson]

The only possible excuse for this is if it's a Precam multi.

 

[Red]

most likely it was imran khan - he had a period (something like 10-15 tests) where he averaged 50+ with the bat and <25 with the ball.

I think this is a more indicative sign of the greatest ARs than individual tests.

Focussing on individual tests is not really a reasonable way to rate an all rounder, because over the course of a series/year an AR can be very influential without ever putting it together in one test.

Over a series we might expect a good batsman (not nec an AR) to fail 1 or 2 times, and it's a bit haphazard as to whether the AR does that in the same test that he dominates with the ball (and depending on the nature of the pitch etc, it may well be likelier).

Just my opinion though…do appreciate the work you've put in...

 

[viriya]

I think this is a more indicative sign of the greatest ARs than individual tests.

Focussing on individual tests is not really a reasonable way to rate an all rounder, because over the course of a series/year an AR can be very influential without ever putting it together in one test.

Over a series we might expect a good batsman (not nec an AR) to fail 1 or 2 times, and it's a bit haphazard as to whether the AR does that in the same test that he dominates with the ball (and depending on the nature of the pitch etc, it may well be likelier).

Just my opinion though…do appreciate the work you've put in...

Good point, it will definitely help Imran's case. He is still just 10% off the top 5 all-rounders though, and a slight bonus based on battingAvg/bowlingAvg (which I'm implementing as we speak), would prob make things more sane. Also, you would expect that issue to affect every all-rounder similarly - just that in Imran's case it was more pronounced.

 

[Valer]

Good point, it will definitely help Imran's case. He is still just 10% off the top 5 all-rounders though, and a slight bonus based on battingAvg/bowlingAvg (which I'm implementing as we speak), would prob make things more sane. Also, you would expect that issue to affect every all-rounder similarly - just that in Imran's case it was more pronounced.

How about some form of closeness factor instead of an overall / per-match split?

Something like

sum(Batting val *sum( bowling value/distance factor)).

where the distance factor is some combination of time and matches played?

 

[viriya]

I added a factor based on the battingAvg/bowlingAvg for all-round careers (~5-10% added points) and reduced the penalty for batting careers of <40 test matches. Changes from before:

Test Batting Career Rankings
George Headley #70 to #55
Graeme Pollock #84 to #70

Test All-Round Career Rankings
Imran Khan #10 to #7
Keith Miller #17 to #14
Trevor Bailey #70 to #64

Thanks for the feedback.

 

[Red]

I added a factor based on the battingAvg/bowlingAvg for all-round careers (~5-10% added points) and reduced the penalty for batting careers of <40 test matches. Changes from before:

Test Batting Career Rankings
George Headley #70 to #55
Graeme Pollock #84 to #70

Test All-Round Career Rankings
Imran Khan #10 to #7
Keith Miller #17 to #14
Trevor Bailey #70 to #64

Thanks for the feedback.

Stick around cricket-web, you might enjoy some of the conversations!

 

[viriya]

Stick around cricket-web, you might enjoy some of the conversations!

Will do.. I'm trying to figure out why cricket forums are so fragmented.. I guess it's too partisan for a global forum to work.

 

[hendrix]

^your posts in this thread are utter garbage. Completely rude and unnecessary.

 

[Red]

^your posts in this thread are utter garbage. Completely rude and unnecessary.

 

[hendrix]

was talking to Lillian Thompson but his posts have been deleted.