Weldone's ODI Rankings

[ankitj]

Read in a bit of detail now. Good work. What I like also is that with normalized scores I can compare bowlers and batsmen too. So Richards, Muralidharan and Tendulkar are top 3 ODI performers in that order. That matches with how I rate top 3 with #2 and #3 swapped.

Beyond that becomes complicated because of all round contributions that can't be ignored.

 

[srbhkshk]

All that I am saying is 35*(40) is same as 35(40). I am not suggesting the 35*(40) batsman should get out. He should carry on and by doing that he's likely to gain more points.
On the other hand, imagine a player who comes to play in the 38th over. Does he do a great job if he ends at 35*(40)? Will you prefer 35*(40) over 25(20) in such scenario? The answer will be no in most cases.

Adding bonus points for remaining not out in successful chases is something I considered before, but dropped that idea for 2 reasons:
1. By the same logic, not outs in unsuccessful run chases should carry penalty, and
2. It will favor players in better teams (Bevan) over others (Andy Flower) because better teams will have more successful run chases anyways.

I do agree that they are basically the same as far as first innings is concerned. But I am finding it hard to wrap my head about this when it comes to chasing. Taking two batsman, one who came in at #3 and scored 40(40) and one who came in at 5 and scored 40*(40) with the team winning, I'd say the second one is a greater innings in at least some way, although certainly not by much. Maybe thinking in terms of resources in a way, a not out batsman in first innings has ran out of resources (same as a batsman who got out in second innings), while the second one (In case any # of deliveries are remaining) has actually ran out of challenges(market?) rather than resources. In a way he overachieved compared to what was expected.

Giving it a bit more thought, I guess penalizing a batsman for staying not out in unsuccessful run chases is also fair, you can not score at the par rate of the match if the situation demands a greater rate. (I'm assuming par rate is constant at a match level?) He used the resources inefficiently and paid for it.

Second point is something that is hard to take into account anyway, will Gilly and Haydos play with the same freedom if they didn't have an ATG middle order and a crushing bowling attack behind them? Probably not. Will Tamim be a better opener if he played for India rather than BD, I'd go out on a limb and say yes. Maybe you can actually have a small penalty for being part of a better team and vice-versa for a poor team?

Anyway, all these are small nit-picks, I do think this is a very accurate ranking is most ways. But I dare say Bevan would climb a bit (and I do think he is lower than he should be), for the aforementioned factor.

 

[smash84]

Weldone, why is Shaun Pollock up so high? and Ambrose as well?

 

[mr_mister]

I love both Ranatunga and De Silva but I can't see how they rank over Bevan. Strike rate not considerably higher and average far lower

I know they both performed in a WC final and all but still

 

[weldone]

Vettori with a higher career ranking than Warne and Morne Morkle higher than Lillee, I knew the Aussies were overrated.

Vettori was such a good white-ball bowler though. Didn't have the SR of Warne & Morne, but was such a banker, & in all conditions, whether home, away or neutral. Had an ER of around 3.9 in his last 10 years.

Warne was also never quite that good in ODIs IIRC. Big game player obviously but it's hard to capture that sort of thing in a statistical exercise.

Yes, nothing much to add beyond what Zinzan and Spark said.

Vettori's consistency of ER playing on mainly green pitches throughout a long career is outstanding. And when he picked some wickets on top that was a massive bonus.

 

[weldone]

Well done, weldone.

The rankings look pretty good in general. Just that i find it hard to put Pollock and Ambrose ahead of McG in ODI rankings (and pollock ahead of Wasim?).

Also Miandad is ranked quite highly, which I approve of. Also good to see Kapil being recognized for his ODI efforts

Thanks for the kind words Smali.

Pollock vs Wasim - Among the 4 pace bowlers you mentioned, Wasim played in the easiest era for bowlers; while Pollock played in the toughest era for bowlers. While Wasim's ER of 3.89 was great for his era, Pollock's ER of 3.67 was at another level for his era. If we measure by this factor alone (ER compared to respective era), Pollock is even ahead of Muralitharan.

Ambrose vs McGrath - This one is more interesting. Both have separate advantages over the other. This one took me quite some time to analyse. Ambrose has slightly better longevity. McGrath was a better wicket-taker in ODIs. Ambrose had better ER but he also played in a slightly easier era. Ambrose had slightly more helpful pitches. All these differences are small. The one factor that makes somewhat significant difference in this case is the quality of opposition batting. That makes sense too because Ambrose played 23% of his matches against the best batting lineup in ODI cricket - the same team that McGrath never had to bowl to.

 

[weldone]

Glad to see that Kapil Dev is a Top 10 bowler.

Also interesting that he is on a batting par with Imran and Watson, and way ahead of Flintoff.

I think bowling and the Imran comparison are both correct (although Imran was easily the better test player). Watson and Flintoff both had short-ish ODI careers in comparison.

In fact he’s ahead of Pieterson, Hayden and Root to name but a few.

Good point, and this is a known issue with our longevity calculation. I have said this to PEWS many times and we don't have a solution for this. In short, our longevity calculation overrates batting longevity of all-rounders. Kapil was a very useful batsman, but to be honest he wouldn't have such a long career if he was just a pure batsman. He was good enough to play as a batsman alone for 6 years maybe, but definitely not 15 years. Ignoring longevity, all the 3 batsmen you mentioned were better than Kapil. Having said that, Pietersen's ODI career ended prematurely and you know why. Root hasn't yet played half the matches he'll end up playing.

 

[weldone]

How did Bevan end up so low?

I love both Ranatunga and De Silva but I can't see how they rank over Bevan. Strike rate not considerably higher and average far lower

I know they both performed in a WC final and all but still

If there's one thing in my ranking that I'm not proud of it is Bevan being at #38. The reason is a combination of his not-so-great SR and my methodology not valuing not outs once the batsman plays more than 30 deliveries.

In my defence, Bevan's batting gets more points than both de Silva and Ranatunga. But they both had twice the longevity of Bevan. Both de Silva and Ranatunga had huge career-spans - from early 80s till late 90s.

 

[smash84]

Thanks for the kind words Smali.

Pollock vs Wasim - Among the 4 pace bowlers you mentioned, Wasim played in the easiest era for bowlers; while Pollock played in the toughest era for bowlers. While Wasim's ER of 3.89 was great for his era, Pollock's ER of 3.67 was at another level for his era. If we measure by this factor alone (ER compared to respective era), Pollock is even ahead of Muralitharan.

Ambrose vs McGrath - This one is more interesting. Both have separate advantages over the other. This one took me quite some time to analyse. Ambrose has slightly better longevity. McGrath was a better wicket-taker in ODIs. Ambrose had better ER but he also played in a slightly easier era. Ambrose had slightly more helpful pitches. All these differences are small. The one factor that makes somewhat significant difference in this case is the quality of opposition batting. That makes sense too because Ambrose played 23% of his matches against the best batting lineup in ODI cricket - the same team that McGrath never had to bowl to.

You raise interesting points although with Wasim he would consistently bowl in the death overs where bowlers would get a pasting. Murali definitely bowled much less at the death and McG usually would bowl 7 of his overs up front (where he would usually inflict early damage). I can't recall much of Pollock for some reason. Almost as if his career went by without generating too many exciting performances.

 

[stephen]

If there's one thing in my ranking that I'm not proud of it is Bevan being at #38. The reason is a combination of his not-so-great SR and my methodology not valuing not outs once the batsman plays more than 30 deliveries.

In my defence, Bevan's batting gets more points than both de Silva and Ranatunga. But they both had twice the longevity of Bevan. Both de Silva and Ranatunga had huge career-spans - from early 80s till late 90s.

Then I think longevity is overvalued.

One thing that does not appear to have been taken into consideration is match situation.

A batsman coming in at 4/60 (15) scoring 4 (2) is inherently worse for the team than a batter coming in at 4/260 (45) and making 4 (2). It seems that the situation punishes both players identically. It is also probably why Bevan isn't as high as he should be. His low strike rate can be attributed to him coming in after a collapse, farming the strike and winning the game. He averaged a strike rate of 80 in the first innings and 60 chasing (from memory) in wins. If you win a game and end up not out batting 4 or 6 chasing them your strike rate doesn't matter at all. Strike rate only matters in a loss or if you get out.

 

[weldone]

Read in a bit of detail now. Good work. What I like also is that with normalized scores I can compare bowlers and batsmen too. So Richards, Muralidharan and Tendulkar are top 3 ODI performers in that order. That matches with how I rate top 3 with #2 and #3 swapped.

Beyond that becomes complicated because of all round contributions that can't be ignored.

Thank you. What's your thought about the Current Ratings? I am aware it isn't perfect for players who played too few matches recently, but does it look like an improvement over ICC rankings?

 

[weldone]

You raise interesting points although with Wasim he would consistently bowl in the death overs where bowlers would get a pasting. Murali definitely bowled much less at the death and McG usually would bowl 7 of his overs up front (where he would usually inflict early damage). I can't recall much of Pollock for some reason. Almost as if his career went by without generating too many exciting performances.

That factor shouldn't matter in my rankings. That's because death bowlers are expected to take more wickets than others (for two reasons: 1. Batsmen slog and have more chances of getting out, 2. They get to bowl to tail-enders more often). And there's high bonus points for getting many wickets in a match. Overall, I think most would agree Wasim's position at #3 isn't too low. All the top 6 bowlers were the cream of ATGs.

 

[weldone]

One thing that does not appear to have been taken into consideration is match situation.

Yes obviously, and no ODI ranking that I know of does that. Ideally cricket ranking should be based on ball-by-ball data and not on match scorecards. That will bring out more nuances and context. That's why I want to base my next ranking (T20) on ball-by-ball data.

 

[smash84]

overall a pretty decent effort

 

[harsh.ag]

This is top work, weldone

Obviously it's not going to be perfect, but then we have our judgment for that.

 

[stephen]

Yes obviously, and no ODI ranking that I know of does that. Ideally cricket ranking should be based on ball-by-ball data and not on match scorecards. That will bring out more nuances and context. That's why I want to base my next ranking (T20) on ball-by-ball data.

What happens, if anything, if you exclude run rate from batsmen who end not out in a chase? Does that elevate Bevan higher in the list?

 

[NUFAN]

Excellent work weldone.

Allan Lamb was surprisingly high to me.

I was wondering was there any tiebreakers for the people with the same ranking score?

 

[Zinzan]

Allan Lamb was surprisingly high to me.

Didn't surprise me. His average of near-on 40 & SR of 75 playing between 1982-1992 is probably equivalent to a high 40s average and SR around 90-95 today.

 

[weldone]

I was wondering was there any tiebreakers for the people with the same ranking score?

No two players ended up with the same points. I have only shown till one or two decimal points, but there are differences later on. For example, you see all of Warne, Starc, Mendis and Kumble have 1.78 career bowling rating points - but there are differences in the 3rd, 4th or 5th (...) decimal points.

 

[weldone]

Allan Lamb was surprisingly high to me.

Didn't surprise me. His average of near-on 40 & SR of 75 playing between 1982-1992 is probably equivalent to a high 40s average and SR around 90-95 today.

I think when I re-run this towards the end of the year, I shall slightly reduce the impact of batting longevity points. That may let AB overtake Sachin and let Dhawan overtake Ganguly but I think overall it will have more positive impacts - for example, Abbas will probably go up, Ranatunga will go down and Bevan will go up in the ranks. I'm not very sure about Lamb without re-looking at the calc but I think he may go down slightly as well.

I don't think I'll touch the bowling longevity effect though - think that one is fine-tuned perfectly.