Thursday 28 December 2023

Thoughts on CTWC2023 finals


CTWC2023 has been over for some time. If you follow the tourney or even my blog posts covering it each year, you'd know it's held somewhere at the bottom of November. Could be a bit later but definitely not after Christmas.

And I completely missed it, only noticing something else which I will also cover here that YouTube started pushing CTWC clips to me. I received zero pushes on CTWC23 clips at all. Yes, zero.

YouTube is of course partially for the blame cause YouTube must have known my interest on CTWC. But then...is it also because the hit rate on classic tetris that is dropping that YouTube decided not to push? But before getting into that let us talk about the tournament itself.

Sadly I don't have time to watch all match ups one by one -- I kind of treasure the time in 2020 where lock down gave me enough time to really go through every single ones. I ended up watching the finals only, so this is my brief thoughts on the finals instead of the whole tourney. Sidnev and Fractals are good representatives of the top circle so my conclusion may apply to those (single digit number of) top players, but not to the larger base like all CTWC finalists.

Very close match between the two. It is very hard for me to tell the difference in style -- the grip may differ but they way they controlled the pieces are too similar. Some luck kicked in and Fractal ended up with a reverse sweep. That requires unparalleled precision but also heart of steel. Good sportsmanship between the two but it had been tradition among the top classic tetris players circles anyway.

...and that's all. Giving further comments is even harder than doing that on the 22 finals where there is at least one unforgettable game that you can talk about. All 5 games in the finals are basically smooth all the way up to Lv.29 then kill screen (or second transition if you like) hits where a roll of dice decides the winner. One may wonder why would it be a roll of dice when the sequence is identical for the two. That is because placement difference is chaotic so eventually the boards would be totally different. Some configurations are weak to a particular sequence and that's the moment that kills.

But wait! Even when you find yourself in trouble due to bad sequence there is still room for rebounce right? Unfortunately we have the Lv.39 hard cap this time. Sounds irrelevant but a hard cap means pure survival is no longer the primal goal in the play. Or rather, a strategy that is relative safe for one to stay at Lv.29 speed over a prolonged period is no longer desirable. Instead you take a high risk high reward route where you would probably die somewhere between Lv.29 and 39 but gives a higher average score (assuming the hard cap at Lv.39) -- if you can't score enough you lose either way hitting the cap or not, so it's better to be aggressive from the beginning.

Last year I said the meta was optimization up to Lv.29 then survive, partly because the cap wasn't there but also because rollers are not super comfortable with Lv.29 speed...yet. This year they got themselves enough space to optimize under such hostile condition, and this is probably what most top players did in recent months.

As the style homogenizes, what is left is basically consistency (ideal input vs actual input), mentality (concentration etc.) and tiniest bit of efficiency. Sure if you count the stats over a bigger sample like the accumulated masters heads to heads you will be able to see the difference among the top players, but in a first to 3 this is just a roll of dice where the better side has a slightly better chance of winning (not to mention lower rounds are just first to 2 sets). (Kind of reminds me about a math question that I just analyzed not long ago -- Simon Marais 2023 B2. I am not going to cover the math here to check my entry if you want to have a look at it.)

This is not by means undermining the credibility of the tournament. A winner is still a winner and I truly believe that every player deserves their final ranking in this tournament. The point is the fun is gone. The commentaries on how close the scores are and how close to top out their players had been are enthralling, but it becomes repetitive soon after some games. The commentators are doing their jobs well, but the game is losing its value to be spectated. 

It might be unfair to compare views count on a fresh video but it is surely dropping, and they might be the reason that they weren't pushed in the first place. It could be a death spiral -- less attention means less new blood, which means less attention in turns. Granted the game has been low key for 20+ years before the current community bought it back, but the spiral is still something one wants to avoid. 

What is the future of the competition, if you ask me? 

There is no reverting in skill and meta progression. Players have proven themselves that Lv.29 speed is no longer a hurdle in general. From this point the remaining progression would be to find the sweet spot balancing aggression and average performance. On the non-capped track the game is even 'cleared' -- Blue Scuti, a very young new comer to the game reaching semifinals this year, survived up to Lv.157 in a game where the game crashes due to memory overflow. Such ending could be viewed as a 'game clear' in the same way as donkey kong or pac-man. I considered the game 'solved' in the sense that all remaining milestones are reachable given enough tries/luck. 

(I am quite surprised that the color shifts didn't cause too much trouble to Blue Scuti though. Other than the really dark blocks at Lv.157 which Blue Scuti almost topped out there are a few more notorious color combinations like dark green and blue at 146 or black and white at 148 but he breezed through the levels fine. Maybe players have practiced enough for that? They recall me of TGM invisible blocks too...)

Classic Tetris will remain as a competitive sport fine despite the lost in visible progression/meta shifts that makes it less appealing. New variations is essential if we want to revitalize the game, but each change will bend it further away from the original game. If it's too far away from NES Tetris people will just as well play modern interpretations like Tetris Effect, so the issue is pretty delicate here.

With top players really comfortable up to Lv.29, I think it will be fine to start from Lv.19 speed just to spare time for later stages. The new Lv.19 phase could last more than 100 lines, somewhere between 100 and 230, but this is up to debate. The time spared can be spent at Lv.29 speed -- actually I prefer a longer Lv.29 speed section like up to Lv.49 considering how quick players topped out with their current aggression. 

Also perhaps a change in game format? Just like how chess is categorized into classic, rapid and blitz we can create multiple rule sets for classic Tetris matches as well. For example the 'uncapped' with no extra cap after Lv.29, a 'rapid' as described above that starts from Lv.19 to Lv.49, or 'first to million' where players optimize to reach million as quick as possible. Some requires slight tweaks to the game but not beyond where modifications have already taken place. The rest are merely change of rules but not changing the game itself. The common theme is to force players into diversified strategy or optimization bias. I expect a much more interesting game in that way. I called a change last year and I would do the same here, and hopefully the drop at anticipation if any, would alert the organizers to evaluate what's going on.

I still look forward to CTWC next year, and if you are interested in the game it's definitely worth a peek into the fantastic world of classic Tetris. It's never too late!

No comments:

Post a Comment