I think we need to start measuring the leaderboard by in game times.

I have tried all different types of emulation to try and replicate the insane load times that the first and second place runs have and I just can't do it. They also don't specify what they are running on, so they shouldn't be valid in the first place.

Hi Anderson san,

Thank you for submitting a run, Its nice to see more interest in Bushido Blade.

The previous runs were done several years earlier, but to my knowledge were not cheated or using unfair means. Since I had joined the board it is my policy to only accept runs with emulation and the version used.

One of the issues with PS1 and PS2 emulation, even on some of the games I have introduced to the site, is that loading can be very different. This game was also released on PSN and Physical PS1 disk, and there is apparently a second print of this disk (not sure what was changed if anything). Not to mention playing the game on PS1, PS2, PS2 Slim, PS3, PS3 slim, and various models. I've noticed those variables can factor into the timing, however I am sorry to say I don't think I could test every variable. As well as the previous runners luck, sometimes the game just seems to load faster in spots, I have gotten stuck on loads before. Emulation can make this issue worse by removing load times, but I don't have an easy resolution, and I don't think its fair to remove the runs, of previous established speedrunners.

The only option I can think to offer, is that if it was desired, you could time the runs, without loads, but you would have to do this for everyone run on the board not just your own. Personally, I have don't think its effective but I have put this out there as a solution in other cases. This is the same response I have given others on PS1 games, because this issue is not exclusive to Bushido Blade.

We can discuss more about it, through message, see my sns options under my name. I think that your run was more than amazing and something to be extremely proud of. I think in other categories you could easily take 1st or have top 3 as well with your approach if thats your goal, or just becoming the best active runner of this game.

Thank you for taking the time to read my response and submit your runs. Hope to see you slash your way through more samurai!

I appreciate your polite response. I took the first place run and my own and removed the load times, which lead to our times being under a second apart. It looks like we might be running / recording at different frame rates based on the timers and how much they advance during each frame of our respective videos.

I will strive to overcome this time gap in some other manner. Thanks for your time!

I have strong ambitions to run this game, in many categories. I also agree that this board, for sake of fairness, should be LTR and not RTA with emulation and mixed consoles sharing the same board. I have 2 runs ready to submit, one on a PS2 slim, one on duckstation 0.1.7836.0. My PS2 slim run was near perfect on the baseline execution, but load times wouldn't let it go below a 1:30. My emulation run, after setting the Read speeds as high as possible, I was able to come close to Intecksss run (around 3-5 seconds), and after setting the scan speeds as high as possible, I accomplished the same jump between RTA times as they had from the prior WR with a 53.817. The overall execution was cleaner in my emulation run as it was more practiced and my focus for getting the fastest possible time to showcase what I was trying to prove versus the feeling of defeat of not being able to come close to a competitive time.

On that note, I did take the time to frame count each fight starting from the first frame the character shows to the first frame of transition to the replay (or the first frame of the death stroke of the final fight), reducing any loading times to almost 0 because the start of the fight replay was fairly consistent and a visual cue that made sense and took cutscenes and mashing to the next fight out of the equation. My initial PS2 slim run was 1162 frames, Intecksss run was 1156, my duckstation run was 1123 and PEPENPAPONTRO's run was 1112. That is from 19.367 seconds to 18.533, well within the realm of plausible, tight times for the fights being as quick as they are. I did these calculations through https://somewes.com/frame-count/

If it is a matter of needing someone to do so and another moderator is needed to retime to LTR, I am willing to offer my time in doing so.

I am a strong believer that emulation runs should show frames, frame rates, inputs, etc. and these are typically required by runs submitted from other emulated systems. While I know that many of these runs are older and not plausible to change this, it should be something considered going forward.

For reference my 2 runs are - PS2 Slim - and Duckstation emulator -

*Edited to show I've been working on the LTR timing so far and each fight frame counts

Bad Ending single character -

JeremyAnderson - vod no longer exists

ConkerXvi - vod no longer exists

267 + 276 + 265 + 261 + 36 = 1105 18.417- PS2 run 53 (my run done live on hardware, not linked above, submitted)

272 + 286 + 266 + 272 + 60 = 1156 19.267 Intecksss

264 + 280 + 264 + 266 + 38 = 1112 18.533 PEPENPAPONTRO

268 + 276 + 272 + 266 + 52 = 1134 18.900 Teru

272 + 284 + 270 + 270 + 50 = 1146 19.100 W0DiE

276 + 271 + 281 + 276 + 50 = 1154 19.233 goatpuncherartic

312 + 384 + 312 + 302 + 86 = 1396 23.667 AonoSaku (movement caused slower time)

276 + 285 + 269 + 276 + 53 = 1159 19.317 JakePlisskenSDA

267 + 284 + 271 + 268 + 40 = 1130 18.833 ayyjer

290 + 308 + 298 + 290 + 82 = 1268 21.133 FunkySteve

284 + 270 + 280 + 264 + 56 = 1154 19.233 GothrockMike

288 + 296 + 290 + 288 + 62 = 1224 20.400 hippokittie

290 + 290 + 278 + 274 + 66 = 1198 19.967 SuperBeastGN

282 + 270 + 478 + 276 + 48 = 1354 22.567 AeronGames

326 + 554 + 326 + 328 + 216 = 1750 29.167 Stuhligen (1:45.400 RTA)

278 + 278 + 280 + 1038 + 62 = 1936 32.267 Radiantstream165

1042 + 342 + 822 + 602 + 176 = 2984 49.733 MacCloud

288 + 736 + 274 + 278 + 844 = 2420 40.333 DBcade (Last hit was hard to tell the exact frame due to blurriness from camera recording screen, so might be off by a few frames)

Normal Ending single character -

PEPENPAPONTRO - Vod no longer exists (would have loved to see it after that Bad Ending run benchmark)

456 + 502 + 588 + 490 + 442 + 752 + 680 + 1472 + 726 = 6108 1:41.800 Stunt_MK

437 + 487 + 603 + 460 + 462 + 1116 + 567 + 502 + 1444 = 6078 1:41.300 JakePlisskenSDA

708 + 537 + 530 + 482 + 576 + 430 + 549 + 478 + 879 = 5169 1:26.150 PS2 run attempt 72 (my run, submitted)

Secret Ending single character -

PEPENPAPONTRO - again vod no longer exists

1000 934 578 1062 950 1150 1506 1438 734 (9352/10524 with area loads) Well jump 470 452 868 696 1522 1068 812 = 15240 / 4:14.000 Stunt_MK (no area loads calculated)

Bad Ending All Characters - (only fight times, calculated as single character Bad Ending, not character/weapon select)

RS - 333 290 268 273 112 (1276) total - 7982 2:13.333 ayyjer

T - 268 328 328 276 100 (1300)

M - 347 340 338 275 50 (1350)

BL - 331 277 341 275 51 (1275)

U - 278 328 273 281 57 (1217)

K - 270 330 283 273 408 (1564)

T - 653 355 267 267 48 (1590) total - 8262 2:16.700 PS2 run 83 (my run, submitted)

RS - 329 282 267 269 459 (1606)

M - 268 712 259 263 36 (1538)

U - 265 278 265 263 27 (1108)

BL - 284 284 281 278 60 (1187)

K - 261 279 363 263 67 (1233)

In terms of overall fairness between the variety of platforms available to run this, I think the Load Removed Time from fight start to fight end would be the fairest way to calculate the runs. I have yet to look at the Slash Mode and Secret Ending times but depending on the area load times of stage 1 for Secret Ending, this could be a weird spot depending on how the platforms handle this differently. I expect to be digging into this over the next few weeks, unless someone already has some more information.

I agree with this opinion.

Basically, emulators and consoles affect loading times. For players who don’t use emulators, it may feel unfair.

I think Load Time Removal (LTR) is fair and reasonable. If not, I would suggest separating the categories into “Emulator” and “Console.”

If the goal is to avoid splitting categories for the development of the game’s community, then LTR would be the ideal solution. For games that require discs, this loading time issue is unavoidable.

I agree. I discussed this prior in the Discord and thought things would be more fair if we used LTR. I did a lot of research because certain emulators like Beetle and Duckstation can get through load screens faster under certain settings while the gameplay remains the same. Normal PS1 copies wouldn’t stand a chance since they are technically the slowest way to play the game. Even PS2 runs will almost always out pace PS1 runs because PS2 had a built in load time enhancer for PS1 games so it will always be faster. I also studied the time it takes for load times on Native PS1 and emulated PS1 runs, and found a few of the runs currently on board that are emulated have slightly accelerated loading speeds than how they should be with the default settings.

With all these variables that can cause loading times to be so different, I think LTR is the way to go.