Sport Today

How on earth can Brian Lara go up 1 point after the Super Series?

lol... really?!!!

What a joke?

maybe its because everyone else have gone backwards quicker

Perhaps he overtook Inzy?

cheers,
Calvin

One point, not one place.

The way the formula works, Lara and company will have received extra credit
because they played against the two top-ranked bowlers (the system doesn't
care whether they *faced* those bowlers or not).  He further benefits from
the low overall scores amassed by his team.

Andrew

actually I think he went up one place because Dravid dropped back 3 places

The OP did say that Lara went up a point, rather than a place. His
points also depend on how the other batsmen in the team perform, I
suppose. Going back to Andrew's comment about benefitting from low
scores amassed by the team, the Bangladesh batsmen should have points
off the charts in that case. I know, I know, Lara did score a quarter
of his team's runs in the second innings.

Here's a plausible theory. The ratings takes into consideration the
last 'n' matches. Probably in his (m-n)th match he sucked even more

Analogy is very similar to Tennis. If you skipped Wimbledon the
previous year, but if you get beaten in the second round in the current
year. You still gain points

or an even clearer example
Lets say Roger Federer skips Wimbledon 2006 and Andy Roddick wins
Wimbledon 2006

Just before Wimbledon 2007, lets say Federer is at 807 points and
Roddick is at 810 points. Federer wins the first round, but crashes in
the second. Roddick wins Wimbledon 2007 too. But due to the ratings
system Federer will gain points and probably #1 even though the results
were totally opposite

OK checked statsguru and around m-15th test mark, Lara had scores of
34 & 6
23 & 0
0 & 8

Also around m-20th test he had
29 & 1 vs Zimbabwe

Surely, if the current scores(vs McWarne vs Aus in Aus) replaces any of
those, you shouldn't be surprised that Lara is up by a mere point

keyboard and brought forth:

The rating simply maps career average onto a logarithmic scale.

It's not the raw career average as you know it, though. For one thing,
the scores in the scorebook are adjusted to take account of how good
the bowling was supposed to be (in rating terms) and whether lots of
other people scored loads of runs or not, and whether the team won,
and then they get weighted as to how recent they are, so that the last
match numbers count at 100%, those from the previous match 99%, and so
on.

So, if Lara's rating went up slightly, that means that his modified
career average went up slightly. 36 out of 144 against the world's top
two bowlers, plus a couple of others who rate in the top 30 or so,
would probably get adjusted to something like 90 and 5 out of 190 to
about 10, meaning that he averaged 50 for the match. If his previous
weighted average was 49.95, it would therefore go up a tiny amount and
he would gain a rating point. Had he averaged 50.05 before, he would
have lost a rating point. (No, I don't know what the exact figures are
for the adjustments they do - those numbers were simply illustrative
to demonstrate the principle of the calculation.)

Cheers,

Mike

 >and then they get weighted as to how recent they are, so that the
last

Agree. But, it also supports my theory that the three consecutive flops
he had now becomes older & less weighted. Since the new score beats
those scores it is not that surprising Lara was up by a point.

36 vs McWarne when thrice the teams couldn't score > 200 definitely
equates to 80 vs Zim or Bangs on a normal pitch and > 100 on a batting
beauty

keyboard and brought forth:

Yes, but his 277, 375 and 400* are less weighted by the same amount.
But yes, if he starts doing better than he was doing recently, his
rating will go up and vice versa.

I dunno about Zim or Bang - it seems worth about 80 against India or
Pakistan, about 180 against the Bangles and about 500 against the now
Streak-less Zims - which is another way of saying that a century
against the Zims is equivalent to about 10 against Australia or 15
against England.

Cheers,

Mike
Cheers

Yeah. Now that I think about it (especially with the Tennis analogy), I
think the model they have come up is pretty good. This ICC ratings and
the DL method is the least appreciated work done by some really fine
mathematicians