Jump to content
HWBOT Community Forums

[OPEN Discussion] Versus OC Game modes


Trouffman
 Share

Recommended Posts

  • Crew

Disclaimer : This thread is aimed to gather opinion and feedback about the Versus OC Game mode experimented by the OCTV Crew. This thread does not predict, ensure or guarantee that your opinion will be implemented in possible formats for fun or as official competitions. Be stay respectful and back your opinion with facts and arguments even if your personal opinion is different.

 

Versus OC is a set of competition format aimed to make benching challenging and entertaining

 

 

It's been in our minds for more than a year an half : How can we extend in a practical way the capabilities of the current 1v1 live overclocking competition format? We've actually conducted some private tests earlier this year under the internal code name of 'Versus OC' but this is the first time that we tested one of the possible formats in live in public. Let us share with you some of the opinion and feedback from this test-session.

 

 

Video replay of the live action 1v1 between Steponz & Marc0053

 

 

 

Overclocking Telemetry Shield Debuts in 1v1v Target Attack Live OC Contest on HWBOT

 

--------------------------------------------------------------------------------

 

OC Games already tested

 

OC Game Experiment #A

 

Suggested Name : 1vs1 Target Attack

Video :

 

Rules are still in the making, refer to this video for the first draft round with a few more explanations :

 

Rules currently (v0.2) are:

 

- Each player has 3 lives per game

- A game is made of multiple sets (max 10), each set = 1 random target (min = top Air Cooling score at HWBOT // max = Top HWBOT Score +1)

- 1 Target is drawn per set

- 1 set lasts for 5 minutes with 1min cooldown period in between each set.

 

Game Rules for Lives

Hit target = end of set, looser looses life, winner gets an extra one (3 lives max per player)

Both No Hit, closest keeps life (delto from player score to target), looser looses one.

Both No score = loose life

Both same score = loose life

Both have same distance = loose life

 

Game ends when someone has no more lives.

Games is maximum 10 sets.

 

 

Feedback :

Perceived Pros :

 

 

Perceived Cons :

 

 

Suggested Improvements :

 

 

 

More details soon

 

 

 

OC Game Experiment #B

 

More details soon

 

 

 

 

--------------------------------------------------------------------------------

 

Useful feedback information

 

To make the best out of the feedback and keep the discussion flowing without being a complete mess.

 

Refer to the Game experiment Tag (the letter + possibly the ID of the run).

And the benchmarks used or try out (number).

 

All constructive feedback with argument is welcome. If you got ideas for rules or mode, drop them here :D

Link to comment
Share on other sites

Game A:

 

- 5 min? Good luck, Asus boards currently (X99 with BDWE microcode) tend to boot longer

- so now instead of ocing and pushing hw to the limit (with the target score) ocers will have to keep pushing cpu let's say 5.1 GHz when it can go to 5.5 GHz easy, so instead of learning pushing hw to the limit, ocers will have to know score range at certain frequencies, used DRAM memory modules, timings and all other crap or tweaks

- tbh "reach the target" is even more stupid than low clock of everything, that's my personal opinion. Pushing to the limits it's true XOC. Pushing lowclock (with proper benchmarks like SPI32M or proper platform - like small sockets for XTU which scale with memories) is tweaking skill & memory pushing and so on, there is some wisdom in it and XOC. Reaching stupid targets in case when HW can go higher is not XOC imho.

- I was saying that 30 minutes it's not enough for World Tour, creating 5 minute different benchmark rounds is imho just horrible, note that OS sometimes needs some changes for specific benchmark

- in theory reaching the target at "easy" frequency is eliminating cpu binning, but what about "lucky run" to get the desired score? Same frequency, same settings and 1 guy gets exact target score and other gets like 1 point less or more, just because benchmarks randomly produce different scores. Example of XTU "742" - same settings and some guys get it after 3rd run of benchmark, with Tapakah we got it after 50 tries of pressing rerun without changing a thing

- again it's livestream event instead of XOC event, sorry but I can't think about many ocers who want to change benchmark each 5 minute, it's just annoying, seriously it looks like "child fun, game" event which we can play during drinking beer and laugh, not for the prizes or places in final of "big live event". I don't see any competitive ocing here

- perfect example of "reach the target" or "predict the score" mess we had in MSI contest for MOA, which still I have no idea how the hell it was judged after failure of the rules :banana:

- you can introduce it for air/water worldtour, makes more sense with limit of the cooling than for XOC in the worst case scenario

- to be fully honest, maybe it would be better if you guys (OCTV) firstly overclocked on LN2 heavily for 1 week all day long trying to get proper scores with few platforms and than start thinking of new formats/ideas for XOC, it's not a personal attack, but seriously most of recent ideas are looking in my eyes as created by a crew without any XOC (especially) live event experience. Maybe after having some "LN2 training" some really good ideas might be introduced. Maybe worth giving a shot.

 

Maybe I am "too old for this s***" but when I see such in my personal opinion mad idea I just want to retire in glory ^^

 

 

Honestly it's even more horrible than current format

Edited by Xtreme Addict
Link to comment
Share on other sites

Time zones and the human need for sleep kept me from watching the trial live, but I've watched the archive now and definitely found it entertaining and engaging, but then people find crazy golf entertaining and engaging - doesn't mean the US open course would be improved by the addition of a miniature windmill.

 

Having been at this for a little under 3 months I obviously have hugely less experience or in-depth knowledge than anyone else in this thread so far, but maybe my perspective as someone who can't be accused of being stuck in their ways is still worthwhile - and my perspective is that I agree with all the above.

 

Firstly, the main thing I noticed watching the event is just how much random chance was involved. The competitors frequently resorted to re-running the benchmark without changing anything, and who can blame them when the output of the benchmark varies by 10x the amount they have to refine their score by? Because of this it feels completely unsuitable for competition - at least anything that points are given out for. I don't think someone who hit a random target by blind luck should end up ahead of someone whose achievements are better but who didn't hit a random target by blind luck.

 

Secondly, in the time I've been at it the main thing I've learnt about overclocking is that it's about hard work and dedication. Many of my best scores (like getting a whole 3.9 hardware points in 3dmark03 with a 210, lol) have been achieved between midnight and 3am after a full day of methodical experimentation and tweaking. Cooling mods, pedestrian though they are, take time to do everything right. Voltmods take time to find, read and understand datasheets and a steady hand for the actual modification. It's not a quick process and not one that can be distilled into 5 minute chunks - or even 1 hour chunks.

 

So, after that criticism, some constructive suggestions - going forward while this format definitely shouldn't be used for 'serious' competitions (in particular anything that gives competition points on hwbot), it could totally work as light entertainment for the OCTV twitch channel. The same element of luck that makes it unsuitable for serious competition makes it more engaging to the casual viewer. Obviously some thought needs to go into what benchmarks are short enough - hwbot prime and ycruncher 25m spring to mind as nice short benchmarks that weren't mentioned in the test run, maybe aquamark as well with good hardware, and short GPUPI runs especially when you consider that there's no need to stick to the standard hwbot presets - but other than that the only real change I'd suggest to the format as a casual competition is using cheaper hardware that is maybe more relevant to the average viewer (IE not LGA2011). It did cross my mind that the same could be said of LN2, but then I think LN2 adds some extra jeopardy that improves the show.

 

As far as broadcasting proper competitive overclocking goes, I don't think cramming the entire process into a twitch-friendly length of time will ever really work. However, maybe a happy balance could be found with an approach a bit like drag racing where all the setup happens off-camera and the only part broadcast is final benching after several hours of setup and test runs? I get the impression - and I'm sure I'll swiftly be told if this is an accurate impression - that all the pros are much much happier having plenty of time to get everything running right and I think the viewing experience watching just the final benching would be just as good if not better than watching someone try to get everything running perfectly within half an hour. If there's a need to add drama to proper competitive overclocking, maybe say once the system is set up, overclocked and has been through test runs and the final benchmarking has started each competitor only gets 5 tries to post the best possible score? But I think the competitive format has to be something the pros are happy won't get in the way of posting the best possible scores - after all, when it comes to proper competitions the possibility of a record being set is much more exciting than onscreen hearts could ever be :)

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...