Introduction
I’ve been starting to learn Guwange lately, and I’ve been switching back and forth between both versions that are available to me (MAME and Xbox 360) on a regular basis to figure out which version I prefer. In doing that, though, I started to notice something else: my scores always seemed much better on the 360. In fact, they seemed much better than they had any right to. Since I’m new to Guwange, I’m still completely garbage at it, but I had exceeded 54万 a few times on stage 1. In contrast, the two arcade counterstop world record replays that I checked (for two different characters) finished the first stage with 57万 and 59万, respectively. Curious, I checked a couple of the high-scoring replays on the 360 leaderboards and saw that one topped 70万 at the end of stage 1, or nearly 20% higher than Yusemi-SWY on his world record counterstop run. What?
Then I noticed that while my personal scores varied significantly between the two platforms, my chain total did not. Following my usual route and assuming I didn’t make any huge mistakes, I would always end up with roughly the same chain size at the end of the first stage regardless of which platform I was playing on, but my score would always still be 20-25% higher on the Xbox 360 version. The same was true of the leaderboard replays I had mentioned: though their stage 1 scores exceeded even the WR superplayers, their chain totals were noticeably lower, as is expected from players that used fewer high-level techniques.
I asked a few people on Twitter to see if there was just a known scoring issue in the 360 version that I hadn’t heard about, but they weren’t aware of anything. I then searched through whatever Japanese resources I could find, and while I found a few old discussion posts alleging the existence of “scoring bugs” in the game, none of them went into specifics, and most of the chatter around the 360 port seemed to revolve around achievements or the arrange mode.
Finally, if you're not familiar with the scoring system in Guwange, for the sake of following along, here's a brief rundown:
Spoiler
In Guwange, your chain total more or less acts as a straight multipler to the value of all of the coins that you pick up; which, in turn, comprise the vast majority of the points that you earn (exceptions being things like static kill bonuses and small bonuses from collecting excess powerups). Basically, during the stages, each coin is worth a value directly proportional to your chain counter, plus an extremely small base value. Depending on what type of coin it is, it may be further divided by 10 or 100. What matters here, though, is that higher chain count = higher score from each coin, in direct proportion. On that level, at least, it’s extremely simple to understand.
Personal Testing
In an attempt to quantify things, I played 3 test runs each with two different characters through the entirety of stage 1 on each platform and compared the results:
The chain averages are roughly the same (actually ever so slightly lower on my 360 runs) but the scores are still dramatically higher on the Xbox 360. I can attest that I was doing the exact same route and hitting basically the same chain milestones throughout the stage in every run, with nothing that would really explain this level of discrepancy. While I don't expect these numbers to actually prove anything, they made me confident enough to actually deep-dive on this to confirm or deny my suspicions.
High-Level Replay Analysis
The next step was to cross-check a number of high-level replays, both from PCB/MAME and Xbox 360, to see if the differences I was seeing applied to much higher-level players, and if it applied throughout the various parts of the game as well. There aren’t a ton of high-level replays on the Xbox 360 leaderboards, but there are a couple above 40 million, so that should be good enough to figure out if there might be something to this or not.
Since most of the stages are pretty nonstop action from start to finish, it’d be almost impossible to try to determine how many points individual coins are worth when there’s lots of stuff going on on-screen. So, I gathered data points from places that met the following criteria:
- No points gained from any other sources immediately before, during, or after the event I was examining.
- No enemies killed immediately before, during, or after.
- No falling coins, homing coins, or leeched coins anywhere on-screen. There is an obvious exception when I'm explicitly comparing the scores from a bullet-cancel that generates homing coins.
- Xbox 360 leaderboards, 46.47m Gensuke replay: mag0908
- Xbox 360 leaderboards, 42.3m Shishin replay: Archvolt (https://www.youtube.com/watch?v=BVhgUaepWco)
- Xbox 360 leaderboards, 36.99m Shishin replay: kenzuuuuu
- Arcade, Shishin counterstop: http://www.nicovideo.jp/watch/sm21915370
- Arcade, Yusemi-SWY Kosame counterstop: https://www.youtube.com/watch?v=J6kdQkLpAs8
- Arcade, Prometheus Kosame 49m replay: https://www.youtube.com/watch?v=M5CKDRe4gR0
- Arcade, Icarus St1&3 reference replays: https://www.youtube.com/watch?v=UTLD8AmJ-Ig, https://www.youtube.com/watch?v=Jhqo1guxzOY
Pointgain from coin piles
Ground-based coin piles are ideal for checking this because their score formula is really simple:
Points = (Current chain counter) + (Coin base value)
The coin base value can change depending on different factors that I'm not entirely familiar with, but I believe the highest value it can possibly be is 9 per pile. Since the base value is only a single digit, the score you get from a coin pile is almost entirely based on the value of your chain. As an additional aside, a coin pile created while your skull gauge is flashing will increase your chain counter by 3 when you pick it up, and if your gauge wasn't flashing, it will instead increase your chain by 2.
I frame-stepped a variety of PCB, MAME, and Xbox 360 replays at various parts of the game and noted the player's score and chain value before and after picking up a large coin pile. Here are the results.
Xbox 360: mag0908 (Gensuke)
Xbox 360: Archvolt (Shishin)
Xbox 360: kenzuuuuu (Shishin)
¹: Chain multiplier as of picking up first pile
²: Chain multiplier as of picking up second pile
Arcade: ? counterstop (Shishin)
¹: Chain multiplier as of picking up first pile
²: Chain multiplier as of picking up second pile
Arcade: Prometheus (Kosame)
¹: Chain multiplier as of picking up first pile
²: Chain multiplier as of picking up second pile
Arcade: Icarus (Kosame)
Conclusion: The Xbox 360 players consistently earned more score from coin piles, far beyond what they should based on the arcade scoring formula.
Pointgain from particular enemies/formations
This is less scientific, but I also observed how many points were gained from certain sections of the game. These generally are enemy formations that are likely to be handled in almost an identical way by all decent players, so the natural variances in score gained should be relatively minor. (The "Chain Increase" column should give you an idea if a section was performed sub-optimally by a given player, though.)
The "Pointgain/Starting Chain" column is a pretty concise representation of the gulf between the platforms. Because the points earned from coin pickups scale linearly with your chain, this value should be roughly the same for all players that execute the same set of actions, have the Skull Gauge filled up high (which, for high-level players, should definitely be the case), and collect the same number of coins in a given situation, assuming there are no differences in the enemy/bullet count.
Stage 1, first arrowhead-bullet turret cart
Note: For this example, because the pointgain is comparatively low, we'll subtract the cart's 640-point destruction bonus from each player's listed pointgain when calculating the ratio, just to keep the numbers more accurate. For all other examples, I won't bother, because the kill bonuses are insignificant.
Stage 1, swarms of bats immediately before the boss
Stage 3, swarms of bats immediately before the boss
Stage 4, first bullet cancel on midboss (from the animal carcass that it spits out)
Conclusion: The Xbox 360 players consistently earned more score throughout the game than their arcade counterparts, especially when weighted by their chain value (which, again, functions as a multiplier and is by far the most significant component in score calculations).
Conclusions
- Points obtained from coins on the Xbox 360 port of Guwange are higher than they should be, for unknown reasons. Additional points are earned from coin piles and bullet-canceled coins alike. I assume the same is true of leeched coins, but those are harder to analyze via frame-stepping since they're generated in large quantities.
- Things like base enemy kill points, boss-kill bonuses, and end-of-game bonuses are correct on Xbox 360.
- If this is all true, on our scoreboard here (and other places like Restart Syndrome) should probably separate 360 and PCB/MAME scores.
- If you're a Guwange player and want to be able to compare your scores to the majority of other players, I strongly recommend playing on MAME.
- I have done no testing for Guwange Blue, but I wouldn't be surprised if similar issues existed there.