First, a small intro - this topic should have been created a year or two ago, but I didn’t think that I would return to speedruns in general, or specifically to speedruns of PlayStation version of NFS High Stakes. Why do I say that I am “returning” - from about mid-2020 to mid-2021, I had a Lap Record on Celtic Ruins. Proof:
1. http://web.archive.org/web/20200930215640/https://www.speedrun.com/user/Tier4Jean (September 30, 2020)
2. http://web.archive.org/web/20210219001810/https://www.speedrun.com/user/Tier4Jean (February 19, 2021)
Video of this ex-Lap Record is still available on my YouTube channel:
The reason why I deleted my run with this video from Speedrun.com - this run became non-competitive due to the fact that at first Shyanji was just a little ahead of me because of advantage that using a manual gearbox gives you. And then ThunderTHR made his Lap Record on Cetlic Ruins MUCH faster than me because of using gamepad stick (I play using a keyboard, and my 1:24.84 Celltic Ruins run was done with automatic transmission)
. . .
And here the transition to the Main Topic begins, Topic which I want to bring up for public discussion
The fact is that ThunderTHR’s last 2 runs (his Lap Records on Celtic Ruins - 1:23.81 , and Landstrasse 1:38.15) were made using PlayStation CPU Overclocking through his emulator settings (DuckStation)
And due to overclocking of PS1 hardware, the in-game stopwatch that measures the current lap time began to work with a TANGIBLE DIFFERENCE from how it works under normal conditions. This Tangible Difference - is difference between in-game lap time and Real Time
First look at the difference between in-game lap time (IGT) and Real Time (RTA) when the game is running under normal conditions (WITHOUT overclocking PS1 hardware)
And I would like to separately note that for an objective comparison I took the runs of several different players, made either on the same tracks as ThunderTHR's "overclocked records", or on tracks with approximately similar lap times. I also specifically chose runs with good lap times, since I don’t see the point in analyzing "slowruns"
1. Celtic Ruins - ThunderTHR 1:26.90 = IGT ( https://www.speedrun.com/nfs4psx/runs/z03nd8jz )
05.848 - start of best lap / start
01:32.835 - finish of best lap / finish
01:32.835 - 05.848 = 1:26.987 = RTA (+0.087)
Further I will not demonstrate calculations, so as not to overload you with information, of which there is already A LOT. However, I assure you that ALL the figures have been checked and double-checked. And if necessary, I can additionally confirm all the data that I provide here
2. Celtic Ruins - Shyanji 1:24.78 IGT , 1:24.867 RTA (+0.087) - https://www.speedrun.com/nfs4psx/runs/zp1qngxz
3. Celtic Ruins - Tier4Jean 1:24.84 IGT , 1:24.950 RTA (+0.110) - video of this run is shown at the beginning of this post
4. Dolphin Cove - ThunderTHR 1:36.93 IGT , 1:36.964 RTA (+0.034) - https://www.speedrun.com/nfs4psx/runs/y2qk5o5y
5. Dolphin Cove - Shyanji 1:32.87 IGT , 1:33.000 RTA (+0.130) - https://www.speedrun.com/nfs4psx/runs/z5ro54dy
6. Kindiak Park - SilkScarfed 1:23.53 IGT , 1:23.583 RTA (+0.053) - https://www.speedrun.com/nfs4psx/runs/yj2pleom
...and now let's look at the difference between in-game lap time and Real Time in ThunderTHR's "overclocked" records:
7. Celtic Ruins - ThunderTHR 1:23.81 IGT , 1:23.566 RTA (-0.244 !! ) - https://www.speedrun.com/nfs4psx/runs/yv1jj4ey
8. Landstrasse - ThunderTHR 1:38.15 IGT , 1:37.881 RTA (-0.269 !! ) - https://www.speedrun.com/nfs4psx/runs/yjg4x9gy
When the game was working normally - Real Time of the best lap was ALWAYS a little slower than the in-game time, even in ThunderTHR's own runs (when he set lap records without overclocking PS1 hardware). But in ThunderTHR's “overclocked records” - Real Time, on the contrary, is faster, significantly faster than in-game time
Even if we take the BIGGEST difference between two runs when the game was working NORMALLY, it will be +0.034 and +0.130 = 0.096
While the MINIMUM difference between a normal game, and a game with overclocked PS1 hardware - will already be -0.244 and +0.34 = 0.278 !! (i.e. 3 times more than 0.096)
And the Biggest difference between a normal game, and a game with overclocked PS1 hardware will be -0.269 + 0.130 = 0.399 !!! (4 times more than 0.096)
Moreover, in addition to the numbers that I gave above, do not forget that in ThunderTHR's "overclocked records” - overclocking PS1 hardware is used, among other things, so that the game produces more fps than it should under normal conditions. Which brings us to the classic question regarding emulators of console games, Is It Even Allowed to use PS1 hardware overclocking (as far as I know, real PS1 console does NOT have this feature) in PSX NFS High Stakes "Best Lap" runs? This is Direct Question to all moderators (Shiinyu, Ewil, goldenONX, wortex), I'm interested to hear their opinion on this matter
. . .
And returning to the numbers and comparison of IGT and RTA timings, which I cited above, I believe that for ThunderTHR's "overclocked records" (Celtic Ruins 1:23.81 and Landstrasse 1:38.15) it is necessary to carry out a special check
This check is very simple - ThunderTHR will have to do at least 1:24.06 lap on Celtic Ruins WITHOUT using overclocking of PS1 hardware. Why Celtic Ruins? Because it is MUCH easier than Landstrasse. Why at least 1:24.06 ? Because on Celtic Ruins +0.25 difference will be quite close to current “overclocked record" 1:23.81, if ThunderTHR fails to repeat or improve his current record
!!! And one more NECESSARY CONDITION - he will also have to show the ENTIRE race from the start of 1st lap (not just a tiny snippet of the previous lap) !!! . This is necessary to prove that ThunderTHR did NOT load emulator save state right before the fastest lap, in order to make his task VERY MUCH easier. Unfortunately, the current "Best Lap" rules literally allow you to "legally" cheat, so in the near future I'm also going to offer a new, significantly improved and expanded version of "Best Lap" rules for PSX NFS High Stakes
P.S. This topic is NOT some kind of "personal attack" on ThunderTHR, or my “revenge" because he broke my old record. So I hope that he, as Experienced and Skilled racer, will pass the check of his “overclocked records”, which I suggested above. And of course ONLY if the conditions specified above are met. Because otherwise (if ThunderTHR refuses / ignores / tries to make excuses / etc.) - suspicions that his Celtic Ruins 1:23.81 and Landstrasse 1:38.15 are Cheating - will increase Very Much
Based on my small analysis of timings (from this thread), in PSX NFS High Stakes "Best Lap" runs - timebug is Not a big deal, if you do NOT overclock PS1 hardware
I even specially calculated the average difference between ALL possible combinations of timings when the game works under NORMAL conditions (0.034 , 0.054 , 0.087 , 0.087 , 0.110 , 0.130) and the result is only 0.043 seconds. In this case (PSX NFS High Stakes "Best Lap" runs) - it is a Very Insignificant number
[ [ [ Calculations:
1.) 53-34=19 87-34=53 87-34=53 110-34=76 130-34=96 (5 timings)
2.) 87-53=34 87-53=34 110-53=57 130-53=77 (4 timings)
3.) 87-87=0 110-87=23 130-87=43 (3 timings)
4.) 110-87=23 130-87=43 (2 timings)
5.) 130-110=20 (1 timing) | 5+4+3+2+1 = 5+5+5 = 15 timings
1.) 19+53+53+76+96=297 2.) 34+34+57+77=202 3.) 0+23+43=66
4.) 23+43=66 5.) 20 | (297+202+66+66+20) : 15 = 43.4 ] ] ]
But what happens in ThunderTHR's Celtic Ruins 1:23.81 and Landstrasse 1:38.15 - I think it should be called not a “timebug”, but “Incorrect operation of in-game time in relation to real time, as a result of interference of third-party software in the operation of the game”. Because in fact ThunderTHR modifies the game, as a result of which game works in a way that it should NOT work under Normal Conditions (more fps, game runs more smoothly)
Once again, average difference between ALL possible combinations of timings when the game works under NORMAL coditions = 0.043
While the MINIMUM difference between a normal game, and a game with overclocked PS1 hardware, will already be -0.244 and +0.34 = 0.278 !! (6.5 times more)
. . .
Summing up everything I said in this topic, I think that PS1 hardware Overclocking through emulator settings SHOULD BE PROHIBITED in PSX NFS High Stakes "Best Lap" runs
Alternatively, everyone just starts using it and we manually retime every run. If everyone is running the game overclocked, everyone is on equal grounds, and thus, not "cheating". If need be, you can filter between real hardware and emulator runs on sr.c, so there is also that if you feel that it'd be unfair towards people running on real hardware (of which there are not many.)
I'm sorry, I should have answered Much Earlier, but unfortunately at the moment I don't have much time and energy
Shiinyu : Alternatively, everyone just starts using it and we manually retime every run
I don’t want to seem impolite, but this “alternative” has neither logic nor common sense. Because using overclocked PS1 hardware using emulator settings will only entail "cons" in the absence of at least one tangible "pros"
The very point of overclocking PS1 hardware is ONLY to achieve a higher framerate from NFS High Stakes in a very controversial way (interfering with the operation of PS1 hardware and making the game work NOT AS IT SHOULD). That is, in essence, a MODIFICATION of the original game, which is generally not allowed, or not recommended by the general rules of Speedrun.com
And of course the "cons" of using overclocking PS1 hardware using the emulator settings:
1. The most obvious thing, PS1 NFS High Stakes becomes MODIFIED, because the Original game should have less (or significantly less) frames per second
2. There is no single standard for “overclocked” settings; in Duckstation alone there is, for example, 300% overclocking 234%, 69%, and so on. And all this Huge number of options (in which in-game time will OF COURSE work differently all the time) will have to be somehow regulated, controlled, checked...
3. It will be necessary to conduct tests on how PS1 NFS High Stakes works when PS1 hardware is overclocked, as well as comparisons with a NORMALLY running game (without “overclocking”). And again, no disrespect to the moderators, but with their current “enthusiasm”, no one will do these tests
. . .
Shiinyu : If everyone is running the game overclocked, everyone is on equal grounds, and thus, not "cheating"
Everyone are Already on equal grounds when the game works normally, without overclocking. Even ThunderTHR himself, when he wasn’t using overclocking, was in these very equal conditions with other players
Again, where is there any logic or common sense in making SIGNIFICANT and Very Questionable changes to the rules - just becausy Only One player (ThunderTHR) in his last 2 runs decided that he could modify the game? Why do I say " Modify” - because as far as I know, neither in the PS1 NFS Hish Stakes, nor in the PlayStation console, you CANNOT go into some settings menu and overclock the hardware to 300%, 234%, 69% and so on
I honestly decided to take my runs down, as I was not aware that the times set in those videos runs could've been affected due to the use of overclocking.
There was obviously no ill intention to try and willingly cheat my way onto the board, it was just a preference thing on how I like to run the game on my own off time.
So is it not possible for the mods to rettime the run or is it strictly not allowed?
Funny how rejecting takes 2 minutes but verifying a legit run takes more than 15 hours. Even with an FPS counter and the emulation speed display.
Luck of the draw in all honestly, plus if you see that a run breaks even one rule, you stop checking it and you move on. The same cannot be said for a valid run, although generally speaking we wouldn't be spending 15hrs on 3min run.
It's way easier to glance quickly over the run and see if it breaks the obvious rules (I do that a lot and it simply saves time watching full runs and then realize that at the end a rule was broken). However if such a run indeed seems fine it will be looked at later in depth. Moderators have up to 21 days to verify runs per site rules so you'll just have to be patient, it won't even take the 21 days or even get close to it honestly with level runs. Patience is the key
