The drivatar, I get better it gets worse apparently!

Forza Motorsport Legacy Motorsport FM5 Discussion
1

This is my first ever sign in to the forums so apologies if this has been mentioned previously, I’m not 100% tuned in to the latest news.
So Forza 5 has been massively promoted on the hype of the drivatar feature, this genius intelligent AI engine that observes and learns how you personally drive. This then takes your ‘skills’ and races a virtual representation of yourself against other players when you’re offline. That’s the hype I’ve read anyway.
However, short of calling !Mod Edit - Profanity - D! on this, as I probably don’t understand the full spectrum of this feature, I’d love an explanation to the following.

I received one of my first ever drivatar rewards on the 8th January 2014. 9 races = 3,870CR. Basic maths will show an average of 430CR per race.
It’s important to note at that stage I’d literally only played as far as the bare minimum to get the drivatar started, I was terrible, didn’t race to any decent standard.
Today, 22nd January 2014 and I’m currently 17th position out of 1,281,106 players on Prague Full Class R Rival. I’ve put in hundreds and hundreds of laps to achieve this and although there are many better than me, that’s a pretty decent achievement. As I witnessed my drivatar percentage training rise, I expected this to be reflected in the dravatar performance. I appreciate this is just one track and one car, well two, but anyone with an ounce of sense could establish an immense jump in skill from then til now.

So I’ve been working away for 4 days, I sign into Forza tonight and I see the following:
Congratulations! You’re drivatar appeared in 170 race(s) and earned 14,250CR since your last session. Basic maths will show an average of 83.8CR per race.
This makes no sense. If this system genuinely emulates how you drive and puts you into real races, since rising from an absolute scrub to being in a top 20, why is my drivatar earning a tiny percentage of what it was at the beginning?
This leads me to wonder if the drivatar rewards are capped or scaled in relation to how long you’re inactive for. Directly showing my drivatar isn’t racing at all, the results and rewards are fabricated.
I also read about you can throw your drivatar into career mode and have it do the races for you. Why? I want to drive, I bought this game to enjoy it myself. Why would I sit back and watch the Xbox drive a car for me. If the drivatar would be competing ‘in the background’ then I refer back to my above results, throwing doubt over whether or not this drivatar feature actually exists or not.
I hope I’m wrong, I hope I’m missing the point entirely.

2

It’s been talked about, but I think all old posts (prior to 1/21/14) seemed to get lost when the new forums went online.

Anyways, I’m not entirely sure what determines your drivatar payout. There was a day where my friend and I looked at ours and our drivatars raced in the exact same number of races for the exact same amount of credits. Never happened before, or since. I wonder if it’s not a little bit random, and has nothing to do with the amount of times your drivatar has actually been seen in races with other people. I raced with my friend’s drivatar one day through 3 different series (including bonus races), and his reward that night said his drivatar was in 5 races.

I’m not even certain if the drivatars you see in a race that are not on your friends list are even real screen names. Nor am I certain if their behavior is accurate and not tailored in some way. By that I mean there are reports that the more aggressive you drive the worse the drivatars are. So the better/cleaner you drive the better/cleaner they do too. But do you really get different drivatars based on whether you race clean or not, or is it just the behavior that is different? Meaning if I am a dirty racer the drivatars I’m against race dirty, but if another player somewhere that races clean happens to have the same drivatar in his races is getting clean racing behavior from them.

So I’m not sure if the drivatars are “this is how so-and-so drives”, or “this is the behavior exhibited by so-and-so”. Meaning the drivatars don’t mimic the same racing line per say, but they are “sloppy”, “aggressive”, “block”, on top of how fast or slow they will be due to the difficulty you selected.

But no, drivatars can not be use to race for you. You could do that in Forza 1. That became the “hire a driver” option in later forza titles, which is not present in Forza 5.

3

Driving one track over and over may give you 100% on THAT track, but not others. Look under My Profile, then My Drivatar. There is OVERALL and a listing of each track/configuration. If the “trained” avatar is good on one track, but hasn’t driven others, it won’t show up very often - certainly less than people who are running Laguna Seca or Sebring’s configurations.

As to the “old” forums, they are available here for a limited time: http://legacy.forums.forza.net/forums/default.aspx If you have any favorite threads of your own, make sure you copy/past the information in the appropriate “new” forums so the information will be preserved.

4

I try my best to race clean, race smart, in career mode. I play on pro or unbeatable, typically finishing 2nd or 1st without scuffing up my car all that much. Most “bumpercars” action is initiated by the A.I.

My drivatar doesn’t seem to be taking notes, though. According to my friends, he still drives like his car has magnets in the bumpers. My sync is at 81% and hasn’t budged in weeks.

5

Apparently my drivatar is a beast in my friends game yet I suck at the game myself lol. I personally call bluff on the whole drivatar thing, more like very heavy influenced AI from the game with a couple of your trates. If the drivatars are supposed to be based on your driving/skill etc… then how can they go from New Driver to Unbeatable with just a quick flick in the settings? Isnt it based on my skill as a driver and what not so how come I can change the settings and now all of a sudden my friends drivatar is a superstar? …not really based on his driving/skill now is it.

6

Firstly, welcome to the forums. Hope you get as much out of the game and the community as I have over the series. Also, congrats on a top 20 time, even relatively early in the game/LBs maturity that’s a good achievement. I’ve not managed one yet. Few top 100s and 50s, but nothing above that. I’ve only been career and a few laps hotlapping to beat friends so far though.

As has previously been suggested, your drivatar should now be skilled on Prague, but its driving on the other courses will not have improved while you’ve not been there. Contrarily, many other people are putting laps in on a variety of circuits, by playing career, rivals or multiplayer, so the other drivatars will be improving on those other circuits. So on those circuits your drivatar relative to the others could well have got worse, and it is entirely possibly that this shows the drivatar system working properly.

I’m not 100% convinced by the system particularly in career mode for many reasons, but I won’t add those to this post. However, I see potential in this system so hopefully it will improve over time.

Jamie

1 Like
7

Ok, so, the counter argument to this is that I’m only training it on certain tracks. And other people are getting… better.

Please allow me to squash this defence, because it’s not good enough.

I’ve just signed into Forza again after 12 hours of being offline, 12 hours after writing the original post.

My drivatar has raced 3 times and returned 3060CR since last night. That’s 1020CR per race!! Last night after 170 races he averaged 83CR? No. I’m sorry but no. This is absolutely nonsense, the drivatar is a lie, the rewards are fabricated, capped and scaled according to your time offline and in no way correspond to a virtual clone of yourself competing in real races.

Do you know what I did last night on Forza, to make my drivatar perform nearly 13 times better than he did over 17 races? I raced 100% purely on Prague, again, for 1 hour.

I wouldn’t make a big deal about this but this is a selling point of the game, and I get really angry at being deceived and conned. And this is a con. Please, someone feel free to explain the differing reward returns and tell me it isn’t.

The only element of doubt that I have is that if this is all a lie, why would it be made so blatantly apparent by this rewards fabrication. It isn’t even negligible or close to realistic.

8

Maybe we need to consider the “unknown” part of the equation, such as the competition level of opponents in which your Drivatar is racing?

Which level is your unknown opponent selecting, New Racer, Inexperienced, Average, Above Average, Highly Skilled, Expert, Pro or Unbeatable? How many laps are they running - is the length of the race, two or three laps or a user-created 20, 30 laps? Is Collision Mode on or off? What level is YOUR Drivatar on any certain track?

It would seem to me, and NOT a lie or a con job, if your Drivatar is showing up at Prague (where you’ve professed a very high ranking) and going against someone attempting to challenge themselves and making more credits by running Expert or above, then you may place higher at the finish. If your opponents, say they’re highly skilled and your Avatar has fallen into their anonymous draw of opponents, is racing on a track where you’ve raced a couple times and haven’t finished particularly well, it may just place lower.

“Differing reward returns” could be a combination of all of the above - choices which the opponents make AND your Drivatar’s particular skills on a given track. The longer the races results in more credits being distributed.

I’m not arguing with you, except about your terms of “lie” and “con” where you’re taking a very narrow view of everything involved.

I like seeing a five-digit payout for a low number of races my Drivatar has participated in, too, but I also realize I’m (1) not as skilled (quick) as you say you are, and I don’t have experience (no time) on all the tracks yet. Maybe I’m really good at Laguna Seca (haven’t honestly raced much there), but am a dud on Sebring. Should I reap rewards equal to Laguna when my Avatar shows up in Sebring. I don’t believe that would be realistic.

9

That’s all fair enough, but let’s discuss the nature of the difficulty settings.
If I race against your drivatar, and I nerf the difficulty level down to effectively blind pensioner skill level… regardless of how good or bad you’ve trained your drivatar on that track, that’s NOT your AI equivalent I’m racing against. If I’m truly racing against legitimately trained drivatars, which difficulty level is the real time accurate representation? Anything other than that exact level is not your trained drivatar, is it an artificially adjusted version? What’s the point in training a drivatar at all when there are multiple skill levels and only one can be the real one? I sincerely doubt the game has been developed to accurately predict and then calculate how you would drive if you were told to drive better or worse, even if it did, I repeat… that’s still not what you’ve trained it to do! That’s all speculation however.
What isn’t speculation is the reward structure. I disagree that I’m taking a narrow view, what I’ve done is studied the drivatar rewards prolonged period of time and presented the results accordingly. I appreciate I’m being overly dramatic by throwing around words such as lie and con, but when faced with evidence like the above and after paying £60 for this game, I’m in no mood to be be PC about it.
To address your last point. Again this is just speculation. I mentioned I’ve only ever raced on Prague, is it realistic to surmise that the 3 races my drivatar competed in last night just happened to coincidentally be on Prague hence why I received big payouts? But over the 170 races, Prague only cropped up a small percentage of times?

I’ve looked back at the rewards, and you can do this too, over the past 4 weeks and the CR reward is each time consistently scaled to the period of inactivity. Regardless of what I’ve done, where I’ve done it.

3 races = 3020CR, 3100CR, 2980CR
169 races = 14070CR, 16090CR etc etc

What I believe is that the developers appreciate that if your drivatar was actually pretty well trained on all tracks, and could realistically come high up in a lot of races, let’s say at 1000CR per race as previous rewards have granted. If I went offline for a week and my drivatar appeared in 500 races… this could potentially yield 500,000CR. This would effectively unbalance and screw up the game economy and intended progression cycle, so they’ve dramatically scaled the CR return down to prevent that from happening. Ergo the reward system is not a direct reflection of your drivatar competing in races.

I know I sound like a stubborn man complaining about something he’s convinced he’s right on, but that’s actually how I feel about it!

10

The various Drivatar settings I previously posted will change the opponents any person races against. If you’re a super star, and they don’t select “super star” as Drivatar opponents, “you” wouldn’t be in the field. The is one of the points I’m making here about your complaint. If I select “Average” for the Drivatars, and your trained Drivatar is “Super Star” you won’t be among those racing.

The assumption you automatically are in each field, should have more races and therefore earning more credits is NOT predicated on “you” but the variables of selection available to all the other players. If you’re a “Pro” all the other players in the game who select “Pro” ratings MAY face your opponent, but it also depends on the track they’ve selected. If they’re racing Sebring, and you’ve never run that track, chances are your Drivatar won’t be in the field.

The assumption that because “your” vehicle should always be active is wrong. What Class do you generally run? If, for instance, you run Class A throughout, but have no cars in B, C or S, “you” won’t be in a field limited to those Classes because your Drivatar doesn’t know them. There is a balance, as I said previously, however a field’s choices of track, Class AND the skill level may exclude “you.”

1 Like
11

I’m sorry, but this just isn’t true.

I drive with “Above Average” Drivatars usually, and I can guarantee at least two friends who are very good drivers will be in my starting grid. If I’m failing miserably at a career race and just want to get gold and move on, I dumb down the Drivatar setting, but those two friends still show up! I have had a number of career races of the “Rivals” category where the game selected these two very good drivers for me to race through traffic against, and I was unable to beat them, no matter what the Drivatar setting. I have literally had to quit the game and restart, until some more beatable friends were selected.

I’ve come to the conclusion the Drivatar rewards are artificial. For example, just now (having played every day for the last three days) my Drivatar netted 3k in 3 races since yesterday’s message. While I was away for ten days on vacation, it netted 13.5k in 169 races. I don’t care, unlike the OP, though, although I wish I could see how well, or badly, my digital doppelganger actually did on particular tracks I like to think I know well.

TTFN
Toni

12

That’s the best explanation i’ve seen for the Drivatar system Snowowl. It’s how i thought it worked, but i wonder as sometimes i do end up with 10 or 12 good racers in my events, and a couple of morons, but i know those people are quite good drivers as they’re on my friends list. I think you’ve got it bang on, but there are some issues which need working on. Sometimes my Drivatar doesn’t update after races etc. These are all things which should improve as the system matures, and we need to all put in some serious wheel time to get the system going, but once it’s up and running, there’s no reason they can’t carry it on to subsequent titles as it’s in the cloud!
Just keep plugging away guys, and maybe a thread with the settings people use might be good. So people could maybe find like minded/able drivers? So people could say, oh i always race C class on pro, or something similar. It might help people to find others to add to their friends list for a better Drivatar experience?

13

I don’t know much about statistics and averaging so maybe my anecdotal evidence won’t be helpful. However, my from experience it appears that the lower the sync the better the avatar. (Granted this maybe different for highly skilled players who start out pretty damn good and become excellent.) This is my first driving game since GT2 on the PS2. And when I started I was flat out bad. Now I’m only mostly not bad. :slight_smile:

I will say this, you cannot judge your drivatar,'s ability by the money he makes. Unless people on your friends list play a lot in career mode it seems to be just dumb luck as to whether or not you’ll even get in a race. There are 14 drivatars just like yours playing rock 'em sock 'em robots through the first four turns. Sometimes it’s just dumb luck who gets out of the pack because of this bumper-car approach by drivatars.

Also there does appear to be rubber-banding in this game. I have had many races where I was > 10 seconds over second place and then all of sudden there is the drivatar of a friend right on my tail. That’s a heck of a pace to catch up at. Granted it could be me driving bad enough to let him catch up, but then you’d think the whole pack would catch up. Nope. Just that one, maybe 2 friend’s drivatars. And it appears to be completely arbitrary which of your friend’s drivatar’s will get the boost.

For the first month of my playing my one friend’s drivatar finished 2nd in every race and occasionally in first. Despite my having 2 friend’s playing and the other friend a much cleaner, better racer. But without fail the lesser played drivatar became Ricky Bobby Trickle Andretti in every race. So I tried an experiment. I created 3 other profiles and did only the first 8 races with them. With one I drove aggressively, smashing for no reason, using cars as turn-assists, and being as dirty a racer as I could. The other 2 I raced as clean as possible for those 8 races. Since it was me driving all 4 profiles (mine and my 3 minions) the driving was the essentially the same.

I should mention that by the time I did this my main’s sync was at 51%.

The toon that all of a sudden drove like he was glued to the pavement was the dirty profile. The dirty drivatar displaced my one friend as the “rubber-banded” toon and became god’s gift to tarmac. Even my other friends had the same experience. Amongst our small group that profile is known as . He’s super toon. And he makes a lot of money for no reason.

My main’s drivatar rarely challenges even though out of my friends I have the best, cleanest times, and I’ve slowly graduated to no assists and manual transmission. I’m generally in the top 10% of times. No great shakes but my point is I’m not awful but perhaps lean towards slightly better than average. On D-class Alps-festival I have 306th best time as of my writing this. Somewhere else, I can’t remember where in Rivals I did it, I have the 83rd best time. I only add that information as justification as to why main profile should reliably place better in my friend’s racing and not as an ego stroke. My zombie profiles were all created with all assists on and automatic transmissions.

My advice, make some from destructo-profiles to annoy your friends and don’t sweat it much. :slight_smile: The new Xbone family policy makes this free and fun. Go into Rivals mode and smoke your friend’s times on every track. You’ll have the added benefit of making a ton of credits. I’m only 1/4 through A-Class and I’ve made 7 million credits and maxed out 9 different affinities. Have fun.

14

Great explination Snowowl. The drivatar system is much more complex than a simple races x performance = money formula. If it was we would have had this a long time ago. It also would not be suprising if with the increase of races your drivatars openent difficulty increases to prevent massive payouts for not playing the game. We won’t ever know that though. All we have to go on is what we see. Like many of you I was a day 1 guy. The first few days the race totals were crazy high and the payouts per race were lower. Now, I hardly see 5 races a day, but my per race payout is much higher. I think this is what Crossy is seeing, but in reverse. The thing is I have a good why my payouts are higher now per race. Most of my friends have finished up career mode and primarily race multiplayer. My drivatar payouts correspond to how much my kids play. They race in “average” and my drivatar will win 90% of those races. Crossy, you probably have the opposite happening. If you are showing up in “unbeatable” your win frequency will be lower.
Think of it as when you started you were the Mini of B-Class and now you’re the Audi #1 Team Joest R18 in P-Class. You’re a whole lot faster, but instead of facing a Mazda RX-8 you have the 787B to deal with.

15

I’ve only quickly ‘scan read’ this thread, but have a similar problem, with other test accounts that I use less actually earning much more.

I think the reason is that once your driver hits 100% training on a particular track, nothing further you do has any impact.

That’s definitely not the way it should work, but would explain what I’m seeing!

16

InfiniteLogic: If you only scan-read the thread, perhaps you’re missing important parts. I would suggest you read through the posts again. Being 100% on a single track is barely touching the surface of “your” Drivatar.

17

OK, having fully read the thread, I still stand by what I said I believe is happening: Once you’ve achieved 100% drivatar training on each individual track (as shown in your driver status), NOTHING further you do on that track (and direction) seems to make any difference whatsoever to your drivatar’s ability and thus payouts. That is born out by my two secondary test accounts and another real account, which I used later after initial experience of driving the early races, all taking part in more races and earning higher payouts than my main account.

I’m very willing to be proven wrong if someone has another explanation, but, ideally, Turn10 should let you completely reset your drivatar and start afresh after you’ve played for a while, or at least take the last 20 or so races on each track as being representative of your drivatar’s current ability rather than just the first few which for me were rubbish as I was playing with a controller rather than a wheel!

18

The one thing I’ve noticed about the Drivatar rewards is that it really doesn’t matter how well I drive, it matters on how much I’ve played in a given amount of time. The longer I play, the more races and credits my Drivatar seems to be in and collect. Just my opinion. It could be wrong and probably is. Cheers!

19

Not sure how this thing is going to work over time but for me it cant be right as the fastest players in my friends list are at the back of all the races i join and these people are like me in the top 300-700 on most of the tracks. Even drivers on my friends list who barely play the game seem to rank higher than a guy who is in the top 200 for most of his tracks. No idea whats going on here, we may just have to see what happens when the thing ha accumulated hundreds of races on all of the tracks

20

as you are getting better so is everyone else. your drivatar isn’t winning as much because more people can beat it. you were specializing in one track when your drivatar has to race every track in career mode. if your drivatar improves it probably goes up in dificulty levels too. the average beginner isn’t going to put the dificulty on unbeatable, they are going to put it on easy. the people your drivatar is racing against now are much better than they were when we all started playing the game.