God of War |OT| Now on PC

Started by Xevross, Apr 11, 2018, 05:21 PM

previous topic - next topic

0 Members and 2 Guests are viewing this topic.

Go Down

the-pi-guy

ScoreNumberRanking
9911
9845
971520
962141
952162
9445107
9358165
9281246
91104350
90131481
89162643
88192835
872171052
862601312
853161628
843451973
833732346
824162762
814523214
805113725
794654190
784964686
775165202
764955697
755086205
744786683
735177200
725177717
715008217
704858702
694359137
684429579
673849963
6638810351
6536410715
6433511050
6331311363
6232211685
6129711982
6023112213
5928012493
5824712740
5722212962
5620413166
5517613342
5416913511
5317113682
5212013802
5113313935
5012514060
4912414184
4810014284
478814372
467814450
457814528
448414612
437814690
424314733
414914782
404614828
394614874
384814922
373214954
363214986
352915015
341715032
331915051
322515076
311015086
301215098
29615104
28915113
27815121
261315134
25915143
24715150
23515155
22215157
21115158
20315161
19215163
18115164
17415168
16115169
15015169
14015169
13015169
12015169
11115170
10015170
9015170
8015170
7015170
6015170
5015170
4015170
3015170
2015170
1015170
0015170







Legend


Xevross


SWORDF1SH


Dr. Pezus


Xevross

10/10 from gameover.gr, will be added to meta.

Man, I swear more than half of the 10s on meta have come from the last 20 reviews.

Xevross

9/10 from 4players.de

That's great! People were worried about this one, as they were one of the least positive on previews and they never give 10/10. The highest they ever give is 9.5/10 and that's super super rare (only ever given to SoTC, rest are 9.2/10 or lower). 9/10 from these guys is pretty much as good as it gets.

So a 9/10 and a 10/10 to add to meta. Should do pretty much nothing to the overall standings.

Dr. Pezus

9/10 from 4players.de

That's great! People were worried about this one, as they were one of the least positive on previews and they never give 10/10. The highest they ever give is 9.5/10 and that's super super rare (only ever given to SoTC, rest are 9.2/10 or lower). 9/10 from these guys is pretty much as good as it gets.

So a 9/10 and a 10/10 to add to meta. Should do pretty much nothing to the overall standings.
Wait didn't they give 10s to a few games?

But yeah that is very good considering the expectations

Xevross

Wait didn't they give 10s to a few games?

But yeah that is very good considering the expectations
Well meta has 100s for some remasters/ DLCs but on their website those reviews actually have no scores. So I think it was one of those cases where meta gives it a score themselves and in 2016 they decided that all these should be 100 even though this place had never given anything more than 92 before.

Xevross

Could someone design a program to solve a large number of simultaneous equations to work out the weighting of each reviewer on metacritic?

Dr. Pezus


the-pi-guy

Could someone design a program to solve a large number of simultaneous equations to work out the weighting of each reviewer on metacritic?
Assuming it's a static weighting, yeah.  It'd be possible.  It'd be a lot of work.  You'd have to scrape all the data from a whole bunch of pages, which would be a project by itself.  Then you'd have to match the data up with variables.  There'd need to be a little bit of leeway to allow things like 94.4 and 94.6.  Stuff like that. 

Assuming it's a dynamic weighting. 
Meta says "the scores are weighted according to the critic's fame, stature, and volume of reviews."  Which it might be:
So you could expect that the first review has less weight than the 100th review from the same reviewer.  Which might have even less weight than the 1000th review.  And you wouldn't want scores to change for older reviews just because the reviewer is worth more now. 

This is still mostly doable, but you'd need to keep track of a few more variables.  Well the easiest one is how many reviews that the reviewer has had so far with each review.  You can probably find a correlation.  It's probably still possible to get the exact value, but it'd be a lot of work. 


Static weighting is probably a fantastic approximation, but there's a chance that it'll be off.

Legend

The best way to do that would be to track the reviews as they come in and see how they affect the placement against other games, at least for the scores beyond the first batch that are added.

This thread just got too technical.
I'm Out!
Go to the hype thread!

Xevross

Assuming it's a static weighting, yeah.  It'd be possible.  It'd be a lot of work.  You'd have to scrape all the data from a whole bunch of pages, which would be a project by itself.  Then you'd have to match the data up with variables.  There'd need to be a little bit of leeway to allow things like 94.4 and 94.6.  Stuff like that. 

Assuming it's a dynamic weighting. 
Meta says "the scores are weighted according to the critic's fame, stature, and volume of reviews."  Which it might be:
So you could expect that the first review has less weight than the 100th review from the same reviewer.  Which might have even less weight than the 1000th review.  And you wouldn't want scores to change for older reviews just because the reviewer is worth more now. 

This is still mostly doable, but you'd need to keep track of a few more variables.  Well the easiest one is how many reviews that the reviewer has had so far with each review.  You can probably find a correlation.  It's probably still possible to get the exact value, but it'd be a lot of work. 


Static weighting is probably a fantastic approximation, but there's a chance that it'll be off.
Yeah that's what I thought. With a static weighting it would be possible but very time consuming.

Anyway, 3 reviews added to meta. The two stated before and a 100 from New York Daily News. It just keeps going up!

Edit: or not, it seems 4players have a big weighting. Its gone down in the rankings.

Dr. Pezus

Yeah that's what I thought. With a static weighting it would be possible but very time consuming.

Anyway, 3 reviews added to meta. The two stated before and a 100 from New York Daily News. It just keeps going up!

Edit: or not, it seems 4players have a big weighting. Its gone down in the rankings.
Harsh critics seem to have a lot of weight

But it's still at 56 for me

Apparently there's another 10 to be added so it'll go net up

Go Up