ICC player ratings relevance

[Vijay.Sharma]

I personally think that the ICC ratings are a good way of determining points per a game because they claim to use different aspects of the game while devising the number. Here's a copy-paste of what they do -

Of course they will never satisfy ever fan's criteria and I guess that is a given. However, on the whole what is your acceptance level of these rating points?

We'll have to live with the fact that they will never let us know the exact algorithm they have used. So we can save time and avoid discussing that - the question is do you think whoever built the algorithms know their stuff or not. Rather, do you believe they know their stuff or not..

Test Match Rankings
For a batsman, the factors are:

Runs scored
Ratings of the opposing bowling attack; the higher the combined ratings of the attack, the more value is given to the batsman’s innings (in proportion)
The level of run-scoring in the match, and the team’s innings total; an innings of 100 runs in a match where all teams scored 500 is worth less than 100 runs in a match where all teams were bowled out for 200. And if a team scores 500 in the first innings and 200 in the second innings, a century in the second innings will get more credit than in the first innings (because the general level of run scoring was higher in the first innings)
Out or not out (a not out innings receives a bonus)
The result. Batsmen who score highly in victories receive a bonus. That bonus will be higher for highly rated opposition teams (i.e. win bonus against the current Australia team is higher than the bonus against Bangladesh.)

For a bowler, the factors are:

Wickets taken and runs conceded
Ratings of the batsmen dismissed (at present, the wicket of Kumar Sangakkara is worth more than that of Makhaya Ntini – but if Ntini's rating improves, the value of his wicket will increase accordingly)
The level of run-scoring in the match; bowling figures of 3-50 in a high-scoring match will boost a bowler’s rating more than the same figures in a low-scoring match
Heavy workload; bowlers who bowl a large number of overs in the match get some credit, even if they take no wickets;
The result. Bowlers who take a lot of wickets in a victory receive a bonus. That bonus will be higher for highly rated opposition teams

 

[OMM!]

They are generally fairly accurate. But they need a bit more consistency factored in to them. It's too easy to slip from 3rd to 23rd in a Test series, which is wrong really.

If you're the third best batsman in the world before the series. A couple of failures don't suddenly make you the 23rd best batsman in the world.

I'd also like them to bring in a wicket-keeper ranking. Seems unfair that the likes of Matt Prior will struggle to make the top 10 as he never bats long enough!

Base it on dismissals, errors, byes, runs scored etc.

 

[GIMH]

They are a decent form guide

 

[Agent Nationaux]

I think that they are fair and that I have never had a problem with them. The only problem I have had is that idiot Pakistani players never excel and improve their rankings.

 

[Vijay.Sharma]

i think that they are fair and that i have never had a problem with them. The only problem i have had is that idiot pakistani players never excel and improve their rankings.

lol!

 

[Zinzan]

They are a decent form guide

^This, nothing more, nothing less

 

[centurymaker]

Yeah it gives quite an accurate reflection of the form of the players.

(though it does have its limitations that we have already discussed on this forum- like awarding more points to players ending up on the winning side)

 

[miscer]

since victories are given bonuses it skews rating in favor of batsmen in good teams but yea it's a good system. edit: then again sometimes you do want to give a batsman a bonus for a win. Like lara's 153*.

 

[Athlai]

Yeah not a perfect system but it goes alright.

 

[benchmark00]

Yeah not a perfect system but it goes alright.

Don't recall anyone asking you itbt.

 

[Athlai]

Don't recall anyone asking you itbt.

As Plops champion I can do as I please ****.

 

[benchmark00]

As Plops champion I can do as I please ****.

Challenge you to a Plops tournament.

Make the thread in OT.



Those are my terms.

 

[Howe_zat]

Think the test rankings make a lot of sense within the top 10 or so, but past that they start to mix up those who haven't played for a while, those who are newcomers, those who are part-timers and those who are a bit **** all together.

There are a few other bugs, mostly that they take the opposition in to account but don't separate the batting and bowling of the opposition - so when India were no.1, they were considered to be the hardest team to score against, which clearly wasn't fair.

The ODI rankings are a lot less meaningful, as stats just don't hold up well in ODIs. Probably the worst thing they do is take economy rate on face value without considering role in the side, or how much the player tends to defend (it's easy to have a low economy when the opposition is chasing 150). Hence the top 10 ODI bowlers are chock full of spinners, three of them from Bangladesh or Zimbabwe.

 

[centurymaker]

im pretty sure they take batting and bowling ratings of the opposition into account too

 

[miscer]

i also hate their knock off 1% of pts for every missed test for whatever reason rule. They shouldn't knock off pts for injuries. Or maybe allow x amount of break but then start taking off pts if the break is huge (because u obviously don't want a batsman rank 1 for 9 months if he's injured and misses like 3 tours or something).

 

[flibbertyjibber]

Was a bit daft that at the start of the summer Bell was ranked 18 despite averaging 90ish over his previous 10-12 matches. Got it up now by keeping that average up against 2 crap attacks. Seems strange really.

 

[morgieb]

It's pretty good IMO. More the team rankings that need to be improved.

 

[Vijay.Sharma]

Think the test rankings make a lot of sense within the top 10 or so, but past that they start to mix up those who haven't played for a while, those who are newcomers, those who are part-timers and those who are a bit **** all together.

There are a few other bugs, mostly that they take the opposition in to account but don't separate the batting and bowling of the opposition - so when India were no.1, they were considered to be the hardest team to score against, which clearly wasn't fair.

The ODI rankings are a lot less meaningful, as stats just don't hold up well in ODIs. Probably the worst thing they do is take economy rate on face value without considering role in the side, or how much the player tends to defend (it's easy to have a low economy when the opposition is chasing 150). Hence the top 10 ODI bowlers are chock full of spinners, three of them from Bangladesh or Zimbabwe.

I think they do consider the ranking of the bowlers as well. I read somewhere a very detailed piece about the algorithm...will look for it and post it in case I find it.

Of course we will never know upto what extent they consider individual bowlers' ratings because that would part of the algorithm and hence their IP. But i am sure they already consider it.

 

[OMM!]

Before the Test series, India had 3 bowlers in the top 10 or 11 in the rankings.

Zaheer, Harbhajan and Ishant were all in there. So any runs in that first Test will have scored bigger ranking points than runs at the Oval!

 

[Howe_zat]

I think they do consider the ranking of the bowlers as well. I read somewhere a very detailed piece about the algorithm...will look for it and post it in case I find it.

Ah, fair enough then. I've been misinformed if so.