Here's a list that shows the current world records that I know of compared to the Non-NBT God standards for each course. I'm using the world records page at the Non-NBT Site to see what the world records are, and I realize it hasn't been updated in a few weeks, so I apologize if there are faster times than the ones at the site. Feel free to fix up the list with faster times, comment about the times, propose a faster God standard, etc. I'm just doing this for fun and to see how the standards compare with the world records. Maybe one day we could make some even faster standards, so that reaching a GOD time would be virtually impossible?

Well, I for one would be willing to make the Non-NBT GOD times harder than they are now. The main reason they were not made "virtually impossible", besides the fact that I thought most players would freak out if they were, is that I planned another ranking system to take over the SR score one anyway at the end, but seeing that the formula I had created for this purpose would remain inaccurate until there would be sufficiently more players inside the rankings, I decided to keep ranking "temporarily" the players according to their SR score.

One main disadvantage the SR score has is that any PR improvement you get is not guaranted to help you get further up in overall ranking and any improvement you make past the GOD standard is guaranteed not to help it.   I can understand that no ranking system whatsoever will be fully accurate, but I nonetheless strongly believe that any improvement, as little as it may be (0"01), should give the player an opportunity to get closer to his nearest overall ranked rival, if it is not to pass him. Well, that is something for which you just can't rely on the SR score system. So basically, I would still maintain firstly, as I did around 1 year ago, that someone's overall ranking should not be solely based off the SR score, and secondly, that the GOD standards should indeed be made harder (especially in a few courses in particular like the KB's and CI1/VL2 in NTSC with the new techniques) in order to really maintain their meaning and keep them relevant and appealing to compete against.

Well, now that this is said, I'm open to any standard change or ranking system suggestion that may come out of someone else's mind, or from mine... I'd post some additional thoughts right away, but I'm not really in a state of mind to do so, and I have limited internet access where I am now, anyway.

Well, my personal opinion on this, to use the very words Chris used, is that the most accurate GOD standards will be times that are virtually impossible to reach.

Obviously, some GOD times, like in the KB courses, have become even easier than the others to be reached over the past few months, but I think if you change a course's GOD time like BC1 to 1'28"99, then most of the other GOD times would also require changes, because BC1 has just been overplayed so much, especially on the PAL half lately, that I don't think you can even compare most of the current Non-NBT WR's with 1'28"99 in BC1 in terms of required 5-lap consistency. The only ones in PAL that come to mind as I write this are the GV2 and MC1 5-lap WR's.

But anyway, I'm not a PAL player, so I could be wrong when judging on the PAL WR's strength. Furthermore, I now really feel like I've set most of the NTSC GOD times slightly easier than the PAL ones, so shame on me for that.   I must say this is annoying me a bit. Courses I've played this year like DP1, BC1, MC2, KB1, MC4, DP3 and KB2 have definitely faster potential than what the GOD times are. Obviously, almost all of these GOD times I haven't reached yet, and I'm not literally saying that they are easy, but what's the use of setting the highest standard promotion to something that would be just barely "moderately challenging"? It just won't stand the test of time, and we'll have to hold this kind of conversation next year again.

The SMK WR's have never stopped to improve over the past few years anyway, and this year was a perfect example of this, so I think we must not be feared to set any eventual GOD standard to something that we might regard as being crazy, "unreachable" at first glance, because we'll know it will be more rewarding in the long run (well, just as long as we don't make any GOD time something that we know will never even get fairly close from being reached (like 11"99 in MC1 PAL)) By the way, I'm not pretending we have perfect knowledge of the course's limits now, and as a matter of fact, I don't even believe there exists any kind of numerical limit in any course (the way I see it is more that the odds for a gifted player to reach x F-lap PR becomes exponentially lower as you go down hundredth by hundredth), but I do believe we know enough so we can set GOD times that will remain challenging for the years to come, unless new and faster Non-NBT techniques jump in along the way. (but then, it's another story altogether...)

Ok, I should have given some concrete examples in my last post. I agree that a GOD time of say 1'00"79 in MC1 would hold practically no meaning, as it would require on average laps within 0"02 from Sami's already insane 12"08. (12"39-12"10-12"10-12"10-12"10) MC1 in PAL is a course which "limits" we've had a good knowledge of for a few years already, so it is unlikely Sami will pull off a 12"05 lap tomorrow or will claim that 12"08 laps have become extremely common inside his 5-lap attempts (even though you could expext anything from him...  ) For this reason, I guess something like 1'00"99 or maybe 1'00"89-1'00"95 for a very fast GOD time would be sufficient because we know that the current WR is already pretty maxed out (but it is still highly probable that a few more hundredths will be cut down off it in the future...)

The same however cannot be said of MC1 in NTSC. That's one place where just cutting a few hundredths off the current WR and calling the result a GOD standard wouldn't be enough if we want one that will really remain challenging regardless of the total time spent on the course. I haven't played MC1 seriously in forever myself, but I know that Sami would get mid-high 11"80's laps without boost on emulator. If you say that an excellent Non-NBT 5-lap could give something around 12"10-11"80-11"80-11"80-11"75, that gives... 59"25. Maybe we could arguably take this into 59"29-59"35 given the difficulty in NTSC to do both the pipe cut AND the last turn boost, but I'm pretty sure really that we just need to practice this course more in NTSC.

To sum it up, I agree with you that even the highest standard should not be set in such a way that it cannot become a realistic goal in itself for any highly dedicated player, but who, even among the highest ranked players, would actually consider 1'00"99 in MC1 PAL to be a realistic goal? The thing is, if we think that a time of 1'00"89-1'00"99 in MC1 PAL is actually fast enough to stand the test of time, then we'll have to go -seemingly- extremely low on all the courses that haven't been maxed out as much as MC1 was in PAL, because chances are they eventually will get maxed out more. Maybe a high 1'15 in DP1 PAL and 1'13"35 in NTSC would do, or whatever we will consider to be humanly doable, but as an "ultimate" goal, not as something to be reached next time the course's WR holder plays the course for a few hours. If someone can improve an already-GOD time by a considerable amount or claim that his average splits would actually give out a faster 5-lap than what the GOD time is, then it is not an accurate GOD time in my opinion. That is, obviously, only if you think that a GOD time should be sort of an ultimate goal, which is what I personally think. (but it is obviously only my opinion; we could give the GOD time any another meaning we want or even re-invent how the whole SR score system works if we truly want to)

About PAL standards, I think BC1 & KB1 could be lowered (1'28"99 or 1'29"09 for BC1, and 0'51"99 for KB1) !

About MC1, I think 1'01"15 is a good GOD standard ! Maybe 1'01"09 is better, but I'm not sure !
The pipejump isn't so difficult, the most difficult is the regularity ! The splits for a 1'01"09 is 12"45-12"16-12"16-12"16-12"16 !! My best 1st lap is 12"43 and my best lap is 12"11 ! And at each race, I did at least one 12"16 !

So OK I could do 1'01"09, and after, I'll push again and again for the unbelievable ! But I'm not sure I have more solid nerves than Sami ! I must confess I'm getting less and less regular, and it's not cos I'm near limits ! It's just psychological !

Well ... put impossible GOD times (like 1'00"89 if I understood Simon), maybe no ! But why not new standards ("ultra-GOD" times, "unbelievable" or "madness" or anything else) with very very very difficult times (like 1'28"75 at BC1, 1'29"99 at BC3, 1'26"99 at RR, ...) ?

But it's only my idea ! And I hope everybody understand my point of view !

Maybe bonus points could be included though?
I am pretty sure non of the current Non NBT GOD standards will get destroyed like they did on Sami's site (so you won't get ridiculous GOD -10 scores).
The standards are too sharp for that

Except KB1 maybe, but even that's within a second

Hmm... this is an interesting system.

The way I was planning to rank the players initially was with a percentage, which formula included a WR total time/personal total time ratio (so, much like your idea), and the player's avg finish. The first conflict back then was to incorporate the avg finish stats somehow with the WR/PR total times ratio in such a way that it still gives a percentage at the end. (and in such a way that someone who has 40 WR's has a percentage of 100%) It did not take me too long to find a way to get around that problem, though the second conflict I had to face was even more existential if you could say so: Give both of the stats a similar weight, and keep the weight balanced regardless of the amount of players there are on the site. That one conflict appeared to be an endless struggle against the numbers, so I ended up just leaving the formula idea under the carpet and go with the SR score "temporarily".

As I said earlier, your idea is interesting though. It converts the WR/PR total times into a score rather than a percentage, and that could help to make the avg finish fit in if we want to, since the avg finish after all also is a total score, only divided by 40. (much like the SR score) The reason I'm insisting so much on the avg finish is that the WR/PR total times ratio alone lacks of human factor. It doesn't take into consideration how much harder going from 1'01"99 to 1'01"49 in MC1 PAL is than going from 1'01"52 to 1'01"02 in GV2 NTSC for example, whereas the avg ranking does it in what you could say is the most "natural" way, by taking the pulse from the players themselves.

Obviously, neither are the WR/PR total times ratio nor the avg finish flawless statistics, but the avg finish generally rewards a lot where the total times does very little (MC1, BC's, RR) and vice versa, so mixing both of them offers a fairly accurate balance at the end to estimate the player's strength, I think. Without total times, people skilled in highly played tracks would be more advantaged than the others, plus any 0"01 improvement would not be granted to help your overall score; without avg finish, some courses would be less rewarding than the others.

Anyway, as I said earlier, mixing both of these stats can be a bit like comparing apples and oranges, so we still have a long way to go if we're aiming for that as a final goal, but so far, we at least have a solid starting point with Chris' scoring system. I have myself messed around with it in my excel files and thanks to copy/pastes, I calculated the total score for every player.   Only, there are 2 things I modified from Chris' formula: First, I only rounded up the final result instead of rounding up all the individual ones (for precision purpose), and second, I divided the time of a WR holder by the 2nd ranked time instead of his own time (the WR), since otherwise, no matter how much he improved his own WR(s), his score would never improve (but as you can guess, that involves total scores going in the negative).

NTSC players:

(5-lap score / F-lap score / Player / Final score)

-8614            -11628            Jamie White      -20241
16061            16062            Simon Laflamme      32123
39030            36484            Martin Morissette      75515
67272            57791            Stanley Jeram      125064
94921            87464            Michael Liem      182385
61538            57886            Christophe Paquin      119424
84647            82928            Juan Sebastián Arévalo      167575
88281            82542            Jason Fraser      170823
122862            110356            Kevin Booth      233218
127558            122069            Ray Bergstrom      249627
39814            42835            Chris Balch      82650
116994            113642            Nathan Stinson      230636
180349            148104            Darcy White      328454
188795            174118            Hue Le      362912
288495            274407            Jonathan Bennett      562902
116985            91057            Jason Whalls      208042
215588            183117            Jean-Sébastien Dubois      398705
210322            174990            Xavier Jutras      385311
284531            234582            Dean Barnhart      519113
89257            90179            Toshi Miura      179436
511864            233404            Drew Blumfield      745268


PAL players:

(5-lap score / F-lap score / Player / Final score)

16318            24687            Michael Jongerius      41006
50642            47948            Nicolas Clementy      98590
44053            49753            Jérôme Gava      93806
46496            58737            Raphael Braun      105233
50429            48346            Gerard Roodhorst      98776
40868            39044            Karel Van Duijvenboden      79912
87919            98833            Serge Ludescher      186752
144686            138282            Cédric Leutwyler      282968
206995            161780            Andreas Runnelid      368775
16006            16292            Pierre L'Hoëst      32298
54239            50533            Matthew McCarthy      104772
-583            -3294            Sami Çetin      -3877
159836            158904            Hampus Lindin      318740
164476            176260            Guillaume Leviach      340736
64053            74389            Armin Huber      138442
92901            113207            Willem van Voorst      206107
26655            25264            Sébastien Holmière      51919
144330            149338            Paolo Galasso      293667
339443            272465            Edzard Wesselink      611907
141325            157009            Dominic Lenz      298334
51182            67812            Matthias Boucher      118994
101030            93267            Franck Mazières      194296

Quote:


Damn, you're so right.   Having a WR holder gain bonus points could have been interesting, but yeah, as you said, it's completely unnecessary. I'm going to modify the formula right away. You know, that's funny, because I really was wondering why the gap between my score and Jamie's was "much" higher than the gap between Martin's and mine, when other stats like SR score, avg finish and total times show otherwise. Now I know why, thank you!  


And regarding the avg finish, I might work on this a bit tonight. I've got some good starting ideas already from back when I was trying to find a similar formula working for an eventual all-MK ranking, but hopefully, I won't end up with a too complicated scoring system.

Player / WR/PR ratio total score (Chris' scoring system)

Jamie White      6591
Simon Laflamme      31081
Martin Morissette      75982
Stanley Jeram      125230
Michael Liem      182385
Christophe Paquin      119514
Juan Sebastián Arévalo      167575
Jason Fraser      170823
Kevin Booth      233218
Ray Bergstrom      249627
Chris Balch      85718
Nathan Stinson      230636
Darcy White      328454
Hue Le      362912
Jonathan Bennett      562902
Jason Whalls      208042
Jean-Sébastien Dubois      398705
Xavier Jutras      385311
Dean Barnhart      519113
Toshi Miura      179436
Drew Blumfield      745268

Avg finish score (actually taken from each individual rank, as explained above) / Player

12016,518            Jamie White
33035,184            Simon Laflamme
66037,392            Martin Morissette
114266,579            Stanley Jeram
176357,293            Michael Liem
120666,926            Christophe Paquin
163752,486            Juan Sebastián Arévalo
169942,012            Jason Fraser
240056,875            Kevin Booth
237916,683            Ray Bergstrom
93426,051            Chris Balch
239982,497            Nathan Stinson
354619,037            Darcy White
348876,623            Hue Le
624444,022            Jonathan Bennett
206654,285            Jason Whalls
399467,623            Jean-Sébastien Dubois
389338,532            Xavier Jutras
511630,956            Dean Barnhart
194772,452            Toshi Miura
642932,273            Drew Blumfield


Final score (the average of the 2)

Rank      Name      Score
1      Jamie White      9304
2      Simon Laflamme      32058
3      Martin Morissette      71010
4      Chris Balch      89572
5      Stanley Jeram      119748
6      Christophe Paquin      120090
7      Juan Sebastián Arévalo      165664
8      Jason Fraser      170382
9      Michael Liem      179371
10      Toshi Miura      187104
11      Jason Whalls      207348
12      Nathan Stinson      235309
13      Kevin Booth      236637
14      Ray Bergstrom      243772
15      Darcy White      341536
16      Hue Le      355894
17      Xavier Jutras      387325
18      Jean-Sébastien Dubois      399086
19      Dean Barnhart      515372
20      Jonathan Bennett      593673
21      Drew Blumfield      694100

Wow, great work there Simon!  I like how you joined the player's rank with the player's PR/WR score. Things are taking shape quite nicely.
Very interesting how some players' overall rank differs between the two scoring systems, yet it's good to know that the players lists look fairly similar heh.
As for comments, well I admit I do like this system. I'll be interested to see how the players lists look after a few weeks of new PRs. What are your plans regarding this system to be used on the Non-NBT site? Personally, I also like the SR scoring system, and the standards are fun to look at and help in motivating me to improve. If you were to use the new scoring system, would you keep the SR scoring system around and use 2 scoring systems?

Again, nice work.

Sorry about the late reply, Simon. No problem about the comments  Now that you mention it, I think you're right that it would be more simple to have one "official" ranking system. It's good to know that the SR ranking and standards will still be used and viewable on the site. It must have taken quite a while to make those standards, eh?
Great idea about archiving the results for each week. I look forward to seeing everyone's improvements. I checked out the Non-NBT site news and it's looking good.

How do you think other players might feel towards this new ranking system?

Hey, I like a lot the new system. Doesn't make such a difference but I think it's better anyway. Right now I can only think of the frustration of having a time 0'01" slower than the standard, which wouldn't exist with the new system.

If you have read Sunday's Non-NBT update already, then you should have understood, unless my writing skills are really that bad, that from next update onwards, the ranking system developped within this thread should start being used to rank players officially. Any comment, suggestion or question will be appreciated, especially given that I haven't received a lot so far. For the time being, I would just like to point out that since Michael sent his new insane-yet-looking-ordinary-to-him MC1 PR's, the upcoming PAL top 3 has switched (back) to this...

1      Sami Çetin      10059
2      Michael Jongerius      38804
3      Pierre L'Hoëst      38954

...Crazy!