1. #1

    [Feedback] Substantial Track Central changes needed

    Skipping over the clearly unfinished Track Central and basic bugs for the moment, there are still some serious issues and features I think need addressing, and addressing early, before too much has happened. Don't worry, I got your tl;dr down below.

    Ratings. The thumb up or down thing simply can not and will not work here. It barely, if at all, functions to distinguish bad tracks from decent tracks, it definitely lacks the ability to distinguish good or great tracks from decent tracks.

    What's more, in order for thumb ratings to have ANY utility at all beyond "probably not an awful track", you rely on substantial groups of users to thumb down tracks that are only decent, and some more users to thumb down even merely good tracks.

    So, for the system to have any slight value, you have builders getting thumbs down piled on their work just for not being great. That hurts, and it will discourage budding builders. Discouraging builders must be avoided with very high priority.

    You just don't allow players enough room in the feedback to do any good for anyone--even for themselves! We need at least the 1-5 star rating, so a good track can be called good, and a great track can be called great. Ratings should be applied by radial menu with "no opinion" in the middle, so that every rating is equally convenient, with always-visible descriptions shown next to all ratings, to calibrate players on the intent of the rating, and with not rating encouraged for anyone who doesn't feel an opinion.

    Display ratings as both a quick normalized aggregate and with a small graph to see the relative weights of each rating. Might be some other helpful display options.

    Naturally, players should always be able to quickly search all their ratings, in order to find old tracks to come back to later and imprint some personal organization on the masses of tracks out there. Players should also be able to see either the 3-5 or 4-5 ratings of other players, and set a feed based on specific players (obviously, you will have a default feed using the friends list) and the tracks they rate highly, to aid in short and long-term discovery. Trials has always particularly let down on long-term discovery and required users to figure out and do a lot of out-of-game work to catch up. When any player can find a few good riders on the leaderboards and then pull in their higher rated tracks, discovery follows easily.

    Ideally, we should have a variety of tags (like the old Technical, Jumps, Thematic, Something Wild, etc) we can set on tracks and then rate 1-5 how that track holds up just on that particular tag. That gives people real tools to find and distinguish the kind of experience they want to have and would be hugely useful, even with only a few users tagging. It also encourages builders whose skills fall into specific areas to get a fair shake and appreciation for what they do well and helps resolve a basic community conflict between tracks that are for riding skill value vs tracks that are for riding entertainment value. No single-rating system will do a good job over such disparate goals, but tag-based ratings perfectly focus on how each track excels.

    Difficulty. Difficulty ratings on tracks continue to be another huge failing on TC, which skews ratings, confuses players, and causes needless strife and consternation. Stop putting it on creators to set track difficulty and let riders vote to determine the difficulty level in addition to the quality rating. Again, a vote menu with some visible notes of common features that distinguish each level from the previous. Evolution's license tests were great for making this clear--Rising is definitely missing that important player calibration aspect, and difficulty levels are completely opaque to new (and probably many old) players.

    Next, periodically adjust the vote weight of riders who don't hit the aggregate difficulty with their vote. That encourages the community to think and act as a hive, which is very important for a meta-feature like difficulty levels. The RL team (probably including trusted community riders/builders) should get a notably heavier weight and even be able to lock a difficulty behind the scenes. That helps calibrate the vote weighting and maintain order, without the community purposely skewing/cheating the difficulty levels.

    Player-voted difficulties also allow you to address the problem of users rating tracks based on the creator's poorly (or accidentally) chosen difficulty, or a poor conception of what that difficulty should feel like.

    It even lets you include Ninja Levels as real sub-difficulty tiers, so the ninja community can both self organize more effectively and present a more accessible front to players climbing into those ranks.

    Medals. Again, it is extremely important to remove medal times from the builder's purview. This requires input from expert riders and an understanding of the community that builders should not be asked to provide. Building, riding, and management are very different skills and goals.

    Anyone who passes a track should receive a bronze medal. Silver to diamond should then scale algorithmically on the times and faults posted to the leaderboard (for example, average a low cluster of zero fault runs to produce the gold target for up to hard level tracks), and they should adjust automatically over time as more times get posted. Platinum and Diamond medals likely should require sufficient competition present to get a good medal spread, as they (diamond especially) are rewards not just for doing well but for standing out. Another detail you can use is how many diamond or platinum medals a rider has, in considering their better-than-gold run in the higher medal calculations.

    Naturally, you will want a quick view where players can see any medal changes they have accumulated, both as simple utility and as a bit of excitement to exploratory riders who want to post early times and see how the community does around them. Allow players to have another view which is a watch list of selected tracks, where they can check their medal standing and see a summary of how other riders are doing on that track. This encourages ride "ownership" so to speak and curiosity on how a track is doing in the community, and how the community is doing on it. A player's friend list (and their watched rider list, which should also be a thing) should automatically feed into this as well, so they can see when they are posting new TBD medals, and join them.

    With proper community-agreed-upon difficulty levels for builders' work, a rating system that actually reflects quality instead of a meaningless up/down, and meaningful medals, you now have the very important ability to tie rewards and accolades to Track Central accomplishments. This shifts the player focus more deeply into TC and what it has to offer as a legitimate part of the game, rather than just a lark or some variety. This feeds back into the system to make TC great and to reward the contributions of builders.

    You will probably want track builders to be able to report glitched or hacked runs from their leaderboards, and it sounds like you need some automated leaderboard cleanup scanners to find new runs with times that don't actually fit the run. Some community driven ways to control cheating on leaderboards will be needed if you can't build the game to prevent it. Better to build a run-verifier that checks from your side as something goes into the leaderboards than to have people needing to police things themselves. A lot less drama that way, and leaderboards and competitions really matter to a lot of people.

    Maintenance. There needs to be a clear way for a builder to mark the chain of creation for tracks, and tracks need to be clearly shown with upload dates and times. It should always be obvious whether one is looking at the most recent version of a track and how one reaches the preferred version of a track, as well as off-shoot tracks for creators who build a track and release multiple versions tuned for different difficulty levels. This valuable practice needlessly consumes significant name and description space ("Fixed bug at CP4" doesn't really inform about the track, and "Final version" doesn't necessarily mean final version).

    We also need proper practice tools, because Track Central and builders' offerings get needlessly cluttered with mini-tracks and segments the like. We should be able to enter a track in "practice mode", where we can set specific points as the "Restart" point, and use buttons to move the current checkpoint forward or backward through the track. Trials is incredibly practice focused, so it is frustrating and bizarre you continue to deny us even the most basic practice capabilities. It also increases the value of user tracks, since players can skip a checkpoint they don't like to better enjoy and appreciate other parts of the track (albeit not to post a run), and it enables building real training tracks, like Trials University has, where the track contains many constructed practice sessions.

    Organization. No general system is really enough for as much as previous Track Centrals have had to offer players, and this problem needs real solutions. Giving players local organizational tools where they can subdivide, create, and name their own lists of tracks is absolutely a necessity. You cannot just leave players paging through flat lists by bare features like difficulty or rating. Build your UI under the assumption that a player is sitting on a pile of a few thousand tracks, and make organization and navigation quick and painless this time around.

    I suggest taking this one more step, too, and allowing players to create their own "stadiums" of TC tracks (maybe even in-game tracks as well) that they can quickly get to from the main track screens (allow a lot more than the 5-10 track stadiums the game uses, of course). Then, allow players to share their stadiums with others, so others can add it in their game and quickly access it and its collected tracks. This adds a curation element to the game and offers players new ways to contribute to the community's enjoyment of your game, plus another natural in-game expansion of discovery.

    Add a "back lot" list for players to privately add tracks to someone else's shared stadium, so they can use the curated stadium as a nexus for their own finds/organization, and allow players to search up common back lot additions to that stadium (seeing what friends have in their back lots might be interesting).

    Activity. Stadium-wide leaderboards will add additional focus points for the competitive types (obviously a friends leaderboard is a must here, as always), and a tournament system could really take the stadium thing to the next level.

    Someone organizing a tournament would, for example, pick a stadium, date range (within limits), submission mode (whether participants designate a run as a submission before or after doing it, and how many submissions you get to attempt or if you just put up the best run you log during the whole window), track weighting (you might want to use a stadium but focus more on some tracks in setting your winner), track bike restrictions (or additions to force people onto an unapproved bike), participant selection (eg, open, added by organizer, qualification round on specific tracks, etc), possible fees, and maybe even some winnings distribution model. Then the tournament system automatically sets up special alternate leaderboards on the stadium just for submitted runs and displays the tournament and results for a period of time. A lot of easily automated stuff here that will be great fun for people to play around with, both in community events and within their various circles of friends.

    Now take that, and let the tournament organizer make specialty tournaments (maybe even just specialty stadiums), where they can apply the contract system to the stadium tracks, so they can pick a bunch of specific conditions as requirements for successful submissions. You already built the framework for this. Might as well let people play around with it and see what they can do. Similarly, let people issue friend challenges by setting up a special contract on a specific track, and then players can pass the challenge back and forth or post it for their friends to run against.

    I suggest polling for ideas on this specifically. Automated tournament management is an easy way for you to hugely support and influence ongoing play.

    Add ways to self organize leagues/teams, and you can foster all kinds of community participation running with no intervention on your part at all. Tons of potential.

    Closing
    I have always thought that Track Central should act as the core of the Trials experience. Built-in tracks make a good base and starting point, but if you build TC right, you could do just as well to throw your tracks in there with the rest and let people go to town (with some guiding hands).

    tl;dr
    • thumb up/down not functional for finding good tracks
    • thumbs negatively incentivize builders
    • need five point quality rating, for players, for builders, for searches, for organization
    • need tags
    • tags carrying ratings ideally supports different kinds of "good" track creations
    • searches using selected players' (higher) ratings
    • track difficulty levels based on consensus
    • consensus difficulty offers more reliable experience
    • consensus difficulty removes "wrong difficulty" rating problems
    • medal targets based on automated algorithmic tiering
    • medal targets not fixed at creation, but adjusted over time
    • players can watch their times change medal tiers
    • meaningful difficulty and medal targets means rewards for TC participation
    • builder tools for track versioning and connections
    • practice mode with selected restart point and switching checkpoints! also, to remove single checkpoint practice tracks and encourage more training/tip tracks
    • manipulable player-named lists to manage large numbers of tracks of interest
    • player-made stadiums to offer quick access
    • shared stadiums for curated track grouping
    • automated tournament handling built from shared stadiums
    • apply contract system to user/group challenges and special tournament rules
     1 people found this helpful
    Share this post

  2. #2
    How come they didnt hire you to build TC? A lot of good stuff there.
    Share this post

  3. #3
    I realy like the idea and world like to add a few additional suggestions / ideas:

    - Improve the performance of track central, the first time I waited over 15 seconds for tacks to load i thought there are no tracks for my choosen filter.
    - For Medals i had the idea of just going by percentages of the leaderboards and taking the "lowest" Person in a "Bracket" as the target time:
    top 1% => Diamond
    2% - 5% => Platin
    5% - 10% => Gold
    11% - 50% => Silver
    51% - 95% => Bronze
    96%+ => No Medal

    So on a Track that got raced by 5000 People...
    Placement 1 - 50 would get Diamong. The time required to get Diamond would be the time that is needed to beat the person on the 50th place.
    51 - 250 would get Platinum time. The time required to get Platinum would be the time that is needed to beat the person on the 250th place.
    And so on...

    It would make the scaling of medals dynamic, making it harder to get a good medal the more good players played the track.
    Share this post

  4. #4
    15 Seconds? Ha! Try one minute on PS4 ... with a 600Mbit/s connection ... and everytime you played a track, it takes that long again ...
    Share this post

  5. #5
    Wow! This feedback is extremely valuable, Redlynx just got this for free. Love and wholeheartedly agree with your feedback, if this was implemented as you suggested it would make a huge leap in the right direction for the game!
    Share this post

  6. #6
    TeriXeri's Avatar Senior Member
    Join Date
    Mar 2014
    Posts
    1,163
    - After playing a track it takes you back to the home screen, so you gotta find the track again if you wanted to add it to Favorites or something, really annoying.

    - Sometimes game gets stuck at bike selection, with no way out except Alt F4, seems to happen when no default bike is selected and other bikes are turned off (black).

    - "Play Later" tracks only go away if you actually finish the track, so if you accidently add a ninja track or something to the list, it's stuck there. Please add an option to remove tracks from it.

    - Like Dislike system doesn't work as well, I'd rather see a system with 5 stars , or multiple categories, sometimes decoration is really good, but driveline really bad, or massive difficulty spikes in design, just having 1 up/nothing/down option says very little.


    - Track Difficulty isn't functioning properly, and not explained in the editor either, so I see a lot of mislabed tracks. And even then, the editor difficulty numbers aren't used. Why are there multiple versions of "easy, medium, hard, extreme" if only 1 is shown.

    - Leaderboards on skillgames that use points or distance, still show up as Time, even while they are sorted on distance/score behind the scenes.
     2 people found this helpful
    Share this post

  7. #7
    Originally Posted by Kadji100 Go to original post
    - For Medals i had the idea of just going by percentages of the leaderboards and taking the "lowest" Person in a "Bracket" as the target time:
    This is definitely a valid way to do it, but Trials has always had medals be personal achievement milestones. I don't think it right to lose that, especially when leaderboard standing is already clear in the leaderboard (though some people would probably like if their leaderboard rank also showed their percentile). At least as far as the normal bronze through platinum, every player should be able to earn their own medals if they work at the track, and the leaderboard is there for competition and comparison. If you tie medals directly to the leaderboard, 90% of players aren't allowed to even get gold, so the net result is to severely reduce what anyone can achieve, and leave climbing leaderboards as the primary play scale, which is definitely not to everyone's taste.

    I was thinking more that we could use leaderboard runs to figure out workable medal targets, rather than just making medals reflect leaderboard rank. As a simple example, the slowest zero fault run is a milestone time. Of course, you can't just use that or people will do super slow zero fault runs to skew the medals, but there's intuitively something "around there", because gold is usually pretty close to a zero fault run that didn't go badly.

    Here is a possible mechanism I've been kicking around:

    First, when runs are submitted, you look at the run and for each checkpoint you take the direct travel time (ie ignore all fault attempts) from the previous checkpoint, along with whether the player was starting there (from a fault) or "rode through" the previous checkpoint.

    Next, we keep just the checkpoint times needed to construct and maintain two reference run times:
    • The "dream reference" is made from adding together the fastest individual checkpoint-to-checkpoint times that anyone has posted, to produce an idealized (and quite possibly impossible) run time.
    • The "lowball reference" is similarly added together from submitted times, but this time we use the best time that started from a fault, to produce a "starting every checkpoint from standstill" time. If there is a checkpoint where no one has started from a fault, we average that checkpoint's time across each of the runs that contribute a from-fault time at other checkpoints for the track (this minimizes necessary upkeep/tracking).

    Probably would hold off setting medals until the reference tracks stabilize a bit and you have a modest pool of runs, but then RL can set medals around those two times to hit the proportions they want (a side effect here is that if people want medals on a track, they need to bring some riders and get the track seen). I think the lowball reference is a solid candidate for silver medal (maybe with a +15% time or some such, since silver medals generally allow several faults), at least for beginner through hard. I still think bronze should just mean "passed the track" as it always has in previous games.

    The lowball reference will need more scaling for some kinds of tracks, especially tracks where faults start to dominate the track experience or people often stop even on good runs. To this end, I think we could track the lowest fault count for each checkpoint segment and when that's non-zero, combine it with that segment's average time-to-fault across our "contributing" runs (ie, one of the runs that sets a best cp-to-cp from-fault time). Might even want to combine total least-faults with the single worst segment's average time-to-fault, since this is supposed to still somewhat lowball a "good run". The lowball reference might be gold or even better on extreme/ninja tracks.

    I might just be overthinking it and you could get just as useful of a result by skipping all the fault/time averaging and using the "best total checkpoint to checkpoint time" (ie, including faults) to construct the lowball reference. That's certainly easy and scales with high fault tracks.

    You would want to play with some numbers here and see how it works out (on built-in tracks, especially), but I think this is enough data to aggregate something useful. I feel like one might want to tease something out of the leaderboard ranking/time where zero-fault runs start consistently appearing. I'd have fun poking around with this if I worked at RL.
    Share this post