Obtaining Better B-Spec Stats ??

[Orion_SR]

I added a couple of columns to your spreadsheet: track #, the target level for that current track, the estimated difference (missing points) and the expected point rise. I also highlighted in green non+13s numbers I expect are normal, and in red if I have some concern.

The points you collected from the Nurb Points Recoup account for the missing points when you began the sweeps. Only 2 more points from Super Speedway is about right, but again - that track is stubborn. It looks like a couple of other points were dropped on the C/World and B/World Courses, but this is not a precise measure.

 

[Famine]

2 more points from A/World (Super Speedway and Infineon Stock)...

 

[Orion_SR]

2 more points from A/World (Super Speedway and Infineon Stock)...

Super Speedway! It figures. I wasn't trying to decode the track names and was relying on number patterns and the average missing points to narrow the focus.

Follow this link to a post that describes what appears to be data from a game where the JP author tried to identify stubborn tracks by placing poorly in the races. Many tracks give up all of the points anyway. Reviewing this page might identify tracks that need special attention when completing the sweeps or looking for stray points.

I'm revisiting your floating points because the +17 for the B-1 Miata is a little confusing. The goal for the Ring was 7051 and 21/29th. At 7034 and ?/29th you were about 17 points short. So I guess it's not impossible to collect a total of 18 points in the Nurb Recoup races. Perhaps if you had reversed the order of the recoup you would have collected a less surprizing +16 with the Miata B, and the expected +2 for the Lupo. This game does some strange things when the fractions resolve - otherwise +14s should be impossible.

Hopefully the resolving fractions can explain the +17 for the B-1 Recoup. Otherwise I need to speculate about bonus points for perfect races (or using brand new cars) that can clean up points from other classes. There is some evidence that machine point bonuses are earned for "perfect races", but there were never any lower-class floating points available when I ran the test; I was quite thourough with each sub-class before proceeding to the next class.

BTW, is there a spellcheck feature for this editor?

D/World/Fu90: A track with +12 is not expected this early in the sweep. Fuji 90s is not identified as a stubborn track, but the average missing points dropped for -17 to -18 after that race. C/World and B/World seem to be missing a point each - on average, but it's difficult to trust the data until the floating points are complete.

 

[Famine]

Just curious what you mean by "perfect race"...

(and one more point just came from C/World/Super Speedway

Edit: I've now retried every 12 pointer twice and am still a point short

Edit eidt: One more point turned up at B/World/Super Speedway, so I'm now at 8098/100/80/80)

 

[Orion_SR]

Just curious what you mean by "perfect race"...

I wish I knew. I was running tests on floating points only. The rules for a perfect floating battle points race are slightly different than for a perfect machine point race. I have no information about how this would apply to course points or track battle points.

Floating Battle: First I collected all of the machine points for a sub-class and then tried collecting all of the floating battle points in a single race. I found that I could if I set a moderarte difficulty level (-2 thru +1), turned on overtake after T2, and adjusted the pace so that the B-driver would use the entire race to slowly overtake each car and finally win the race. Or, if he passed the leader after one lap, I could pit and allow a large lead, make the come-back win, and also collect all the points. The perfect race for F-battle is very similar to mission 34. The strategy that works the best is trying to keep that little orange arrow indicating a faster split in front of the car as much as possible. I have some amount of control of this strategy and can apply it to other cars and races, and at different difficulty levels.

The perfect race for machine points is beyond my control. The JP author seems to have mastered this skill and judging by his scores, it is more of a priority than all of the F-battle points in a single race. I was trying to match the floating points scores of the JP author and found it very difficult to collect as many machine points as he did on the second or third attempt with the same car. He was pulling in 32 and I was lucky to collect 11 and every once in a while collected 22 - a little more than twice as many as would be expected by practicing for two laps. So I kept trying different line-ups and paces and finally pulled in a 30 something score.

I eventually had fairly regular success with the single bonus (22 instead of 11).But only saw the double bonus a few times in many attempts. Dramatic come-back wins seemed to be a factor for these races, but maximizing the sector arrow wasn't helping at all. This was the first indication that perfect races could pay of with large bonuses. There are clues to win bonuses in the JP web site, but I'm not sure if they are related.

BTW, for both machine and F-battle points, following cars and not overtaking caused a significant reduction in the points earned each race. Passing, pitting and catching up is better than follow the leader.

 

[Famine]

Gah!

I got up to 8098 (see last edit) and chugged on to B/City Courses...

Race 87, should have 8125, got 8124... At 8201 neither Course nor Battle ticked over to 82 (until the next race) and at 8304 only Course ticked over to 83 (leaving Battle until the next race). As far as I can tell, this means I'm still missing a point - or rather 2/29ths of one - somewhere, probably in D/World Circuits somewhere.

Bugger bugger bugger bugger bugger bugger!

 

[Orion_SR]

Bugger bugger bugger bugger bugger bugger!

I agree. Don't worry about the course skill lagging a point - that's to be expected since the skills seem to need whole points. For instance, the skills don't change at 7400. Also, the battle skill lag tends to be about 3 times the missing points. If you continue the sweeps without tracking down the missing point, you can try monitoring this lag to make sure it doesn't creep higher.

 

[Famine]

Out of interest, have you ever known a track to give up no points when there are actually points left to get from it?

Edit: I think I'm now 2 points down

 

[Orion_SR]

Out of interest, have you ever known a track to give up no points when there are actually points left to get from it?

Edit: I think I'm now 2 points down

Yes. Super Speedway is infamous for holding on to points. Looking through the JP Stubborn Track data I see that happened on Deep Forest reverse, and Autumn Mini forward and reverse with the 350Z.

I tried comparing your records to the JP Past Relic collection. When the JP author dropped a small fraction, very little change occurred to the collection pattern. Only the milestone scores came up a little short. I would not have expected a +12 on the race after the 8125 milestone unless you were short almost a whole point, and not a small fraction. Another +12 only 6 tracks later is also unexpected. In this case the +12 track may not be the problem. A fraction could have been dropped on any of the preceeding tracks forcing the +12 to occur early. The 3rd +12 doesn't show up until after 19 more +13 tracks. I would have expected it a little sooner.

Are you using R3 tires on the Miata + SC? You could try a more powerful B class car. That Miata + SC is at the lowest end of the class, but I haven't tested track points to see if this makes any difference. Still, it seems logical to use the alternate tuning the second time over the same track.

 

[Orion_SR]

Please let me know if any of the more recent information or strategies are particularly useful. As I re-read many of the older messeges, I feel the need to re-write a lot of it in light of more recent information. At the very least I need to gray out incorrect statements. However, I'm a little reluctant to make major revisions to such old posts. Perhaps it is time for a complete revision under a new topic. I can't even find the messages I need to reference anymore.

What I have in mind is similar in structure to the GTASA/PC record book thread. When I built the topic I ran into problems with spam filters and other user limits as I reserved the posts I needed. I'm not sure what advanced options might be available for managing this kind of topic. Also, I'm not sure how this would support or conflict with Sportwagon's reference page. Actually, I probably won't have much time to work on this until June.

 

[Famine]

If you make out an entire new thread containing everything you know about B-Spec, I WILL sticky it.

 

[SportWagon]

I can add the new link to my reference page, perhaps demoting other links to "for historical interest only", or something like that.

I'm busy right now, but I'll reread my first post in that thread with a view to incorporating what you put in later posts.

Etc.

Okay, I copied your previous additions into the first post (although the thread was really short enough that arguably that's redundant. I'm having a hard time "throwing out" the old favourites, however.

 

[Orion_SR]

I updated my post on the reference thread to include links to the more recent messages. I wanted to add a link to a post that describes the influence of HP, Tires, Downforce, and Weight on A-spec points as a guide to tuning cars to meet sub-class requirements, but I can't find the link to that chart I was using. I think it was in the less favored A-spec thread. Can either of you reference a post that might be useful for this purpose?

BTW, SportWagon. Your last post on the reference thread is now totally out of context. But yes, the software translators seem to be doing some phonetic translations. I was trying to translate some of the cars referenced by the JP author an noticed a similar pattern. I found it interesting that software seemed to have problems mixing up Ls and Rs. Why would software be influenced by a language stereotype?

We can't throw out the old messages yet. Most of my posts assume the basics were learned from those message and are incomplete by themselves.

 

[SportWagon]

I found it interesting that software seemed to have problems mixing up Ls and Rs. Why would software be influenced by a language stereotype.

Someone with better knowledge of Japanese may want to correct me, but as far as I know, there are no two distinct sounds "L" and "R" in Japanese itself, so if what they're doing is transliterating foreign words into readable Japanese, the same Japanese word will do for both cases. I.e. I think their closest natural consonant is actually somewhere in between the two.

So a Japanese word used to represent the foreign word "Race" may in other cases represent the foreign word "Lace".

So the translator knows the particular word (symbol) is frequently used to phonetically represent the foreign word "Race" or the foreign word "Lace", but really can't tell which.

I wanted to add a link to a post that describes the influence of HP, Tires, Downforce, and Weight on A-spec points as a guide to tuning cars to meet sub-class requirements, but I can't find the link to that chart I was using. I think it was in the less favored A-spec thread. Can either of you reference a post that might be useful for this purpose?

https://www.gtplanet.net/forum/showthread.php?t=70995 ?

 

[Famine]

There isn't actually any "L" sound in Japanese (kinda - depends which Romaji method you use really), so any L or R sound is written as an R. This makes turning the word back into English quite a pain, as an R can be an L or an R. You'll notice that in Grand Valley (gu-ra-n-do ba-r-ri), both occurances of the r-based syllable are turned into l by the translator (Glan valley).

The R is pronounced with the tip of tongue against the back of your top front teeth - exactly the same position you'd take if you were going to use a Western D.

If you want to hear a Japanese person committing seppuku, get them to say "cylindrical" (sp. shi-ri-n-do-ri-ka-r-ru) or "orange-tipped fritillary" (sp... who knows?).

 

[SportWagon]

If you want to hear a Japanese person committing seppuku, get them to say "cylindrical" (sp. shi-ri-n-do-ri-ka-r-ru) or "orange-tipped fritillary" (sp... who knows?).

That's a flower, not a butterfly, I'd guess?

Hmm. No. Web seems to favour butterfly.

 

[Orion_SR]

Estimating B-spec Sub-Classes Based on A-spec Car Values

Comparing Wild Cobra Z28's Car Value data with the Family Cup Opposition Table and JP prize car data, I have a reasonable chance of calculating the sub-class value range for B-spec points. Initial estimates show a fairly predictable increase of a value of about 140 for each sub-class.

A-1: 0824 2002 VW Lupo 1.4 (or less)
A-2: 0946 1980 Renault 5 Turbo
B-1: 1113 1971 Dodge Charger Super Bee
B-2: 1249 2003 Mazda BP Falken RX-7 (D1GP)
C-1: 1375 1986 Peugeot 205 Turbo 16 Evolution2 Rally Car
C-2: 1511 2001 Mazda RX-8 Concept LM Race Car
D-1: 1650 1997 BMW McLaren F1 GTR Race Car
D-2: 1745 1998 Suzuki Escudo Dirt Trail Car (or more)

Review Wild Cobra Z28's thread A-Spec Point Data; Values, Races, and Modifications for information on calculating how changing the power, tires, weight, and downforce will influence the value of a car. This value appears to correspond to which class or sub-class of B-spec points will be earned.

 

[Orion_SR]

Oops, sorry. I duplicated my post above.

 

[Dimitrov]

There isn't actually any "L" sound in Japanese (kinda - depends which Romaji method you use really), so any L or R sound is written as an R. This makes turning the word back into English quite a pain, as an R can be an L or an R. You'll notice that in Grand Valley (gu-ra-n-do ba-r-ri), both occurances of the r-based syllable are turned into l by the translator (Glan valley).

The R is pronounced with the tip of tongue against the back of your top front teeth - exactly the same position you'd take if you were going to use a Western D.

If you want to hear a Japanese person committing seppuku, get them to say "cylindrical" (sp. shi-ri-n-do-ri-ka-r-ru) or "orange-tipped fritillary" (sp... who knows?).

That's why is so much fun speaking english - or your local language - with nipponic syllabs.

"Let´s race in Nurburgring with our VW Golf"

"Japanifying" try: Retuso reso in Nuruburugurinugo wiso auaru borukusubagenu gorufo.

 

[Famine]

I should just like to add "GAH!":

 

[Orion_SR]

I'm looking at your data set, and I've add a few columns to help me see the number patterns. N is the track number (5-232), O is the target level, and P is the difference between your level and the target. This reveals a pattern of -1s and -2s similar the the pattern of +12s and +13s.

The pattern for the last cycle, tracks 204 to 232, is the same as the previous cycle (175 to 203). This would indicate that nothing changed in last cycle, so let's not worry about those tracks for now.

The pattern for cycle 146 to 174 doesn't match cycle 175 to 203. The +12 occurs on track 186 instead of 187. This would seem to indicate that a tiny fraction was dropped somewhere between track 158 (row 236) and track 186 (row 268). The most suspicious looking tracks are the last few completed with the RSC - Autumn Ring mini through Beginner. Several of those tracks tend to be a little stubbon (maybe High Speed Ring, rows 241, 242, 267, and 268). This point is likely to be too small to be detected when collected unless it's the very last point.

The pattern for cycle 117 to 145 matches the pattern for cycle 175 to 203 and this seems to extend back to track 100 (row 172). These track seem to be complete: 100 to 158, and 187 to 232.

On track 86 you were are at -0 but missed the milestone at 8125 by -1. Something is missing before then. I suspect this is a fairly large fraction because you still earned +12 on the next race and stay at -1 for the next few races. Then on track 94 you drop to -2, so points have definately been lost somewhere between track 87 and 94. Also, the data for 95 to 99 is useless - I can't tell anything with the small data set.

It looks like you were quite thourough in your A-D/World Again search. Earning points on InSt and TRMS with the MX-5 explains most of my suspicious data. The only number that might be out of place is the +12 on HoKR with the Lupo - maybe recheck tracks 61 to 73 with the MX-5.

Best Guess Summary

Large fraction before track 87 (8125)
* Maybe the MX-5 (A), tracks 61 to 73, rows 121 to 135
* Points might be missing on more than 1 track

Large fraction between tracks 87 and 100, rows 159 and 166
* MX-5 SC (B), maybe recheck tracks 95 to 100
* Points might be missing on more than 1 track

Small fraction between track 158 and 186, row 236 and 268.
* RSC Driving Park 1st, then Original Circuits,
* Also HSR+R with MX-5 SC and Driving Park with 350Z LM

For clarification: I cannot apply math to the fractions to be more precise. Strange things happen when the fractions resolve which makes estimating the size of a fraction and predicting +12 very difficult.

It looks like you are missing 2 points on at least 3 tracks, so earned fractions may be hidden. Chart and save all the +0 tracks. It's up to you to consider resetting after a +1 track in order to determine which tracks were actually hiding the points.

 

[Uncle Harry]

My B spec search started about a week ago.
I started at 7148 / 87 /70/70 from racing Enduros, F1 and other long events.
Currently now at 9600, 100 machine, 96 course and 95 battle.
I have used a Mini, a second hand RX7 stock at 206 hp and the Toyota Rally Raid race car at all family cup races.
Most of the remaining 400 points I hope to find in the 58 family cup races using class D cars.
It looks like I am lagging a bit in my battle score.

If your score goes up by 100 points does it mean I picked up floating points?
I used a stock power mini (90 hp) with transmission and suspension mods. In 8 races I picked up 12 machine skill points (87 to 99) and my total B spec jumped approx 800. I did not make notes but I was paying attention to the score after each event.

Machine skill hit 100 after 20 more races. (does that mean I have all floating points).

Interesting 18 page read.

 

[Orion_SR]

Any Mini should be safely within the A class and the RX-7s look like good choices for the class B track points. And the RSC is known for collecting class C track points, so you should be good to go for class D. However, it appears that you skipped the Nurburgring tunings so there is some concern about the floating points.

Fresh tracks should only give up +13 points - maybe +12 or +14 depending on the fractions. So yes, if you earn +15 or more then you are definately collecting floating points. If you keep earning points by repeating a race you are probably collecting floating points.

A stock Mini (One?) at 90HP should be in the A-1 sub-class. Many points in this lowest sub-class tends to get missed because of upgrades and very few races can be won with an A-1 car.

A machine skill of 100 means you have all of the machine points, but it does not confirm that all floating battle points are collected. The rules for earning floating battle points are different than the rules for winning endurance races using B-spec. However, machine points usually take longer to collect than the F-battle points.

There are 8 sub-classes of cars for earning all the floating points. I always recommend collecting the floating points before track points so they don't get mixed together and confuse the data. However, I suspect you will want to finish the D class track points and worry about floating points later. Since you aren't keeping notes there won't be any data to confuse, so this plan is as good as another.

Battle skill is usally the one that lags. If you had more data I might be able to tell you more about the type of your missing points, or maybe how many are missing, but this doesn't usually help you find them.

 

[Orion_SR]

Interesting 18 page read.

You actually read the entire topic? I hope you aren't more confused than when you started - there is a lot of misinformation in this thread that gets worse with age. I need to clean things up in a new thread. In the mean time, check the following thread for direct links to many of the key messages.

B-spec Point Maximizing Quick-Reference Links
https://www.gtplanet.net/forum/showthread.php?t=69405

 

[Uncle Harry]

I pushed my score up to 9824, 100, 98, 97 using the Nissan 350 LM with most of the city tracks left. I dont think I have to many more tracks that I have not used in B spec before with this class of car.
So I think you are correct in assuming I need to take the stock Lupo out for floating points at the ring.
I may also have to modify a class b car for more floating points.

I read all 18 pages and the other main links as well.
Thanks for your reply Orion_SR.

 

[Orion_SR]

If you used a Mini One, then the A-1 sub-class (Lupo) is probably done, but there still might be some floating battle points, so it's worth a shot on Nurburgring. The Mini Cooper is just above the level where I would expect it to earn A-2 instead of A-1. The Mini Marcos and Mini Cooper S should be well within the A-2 sub-class. So if you were using the Mini One, you might have better luck with the Miata 1800 RS, otherwise the Lupo, or better yet a Nissan Micra, should clean up any missing floating battle points.

An RX-7 looks like a great choice for sub-class B-1, but that still leaves B-2 unchecked. There are also alternate sub-class tunings for your RSC and 350Z LM, but since these are JP method cars you shouldn't have a problem identifying the sub-class used.

 

[david_number1]

Ive got 8715/99/87/86 B-spec points and am wondering what to do next? I havent tried to increase them so should i start at the beginning of the JP method or what? Any help is appreciated!

Cheers

 

[Orion_SR]

Yes, always start at the beginning with the 8 tunings on Nurburgring before working on the track point sweeps. Since your machine points are not maxed there are still some floating points available. My guess is you will probably max your machine with an A-1 car (Lupo, Micra), but it's a good idea to try all 8 tunings to clean up any stray floating battle points.

Or, don't bother. You have plenty of points to win any race using B-spec, the driver won't get all that much better with more points. Collecting all 10000 points is a long and frustrating process that frequently ends with a few missing points anyway.

 

[Uncle Harry]

Week end update.
9938, 100, 99, 99.
Just to concur with Orion_SR. Very frustating and my B spec driver still stuffs up.
Do I have to find 4 or 5 races where I overtook to quickly from a standing start?
What is the 2nd tune level for the RSC Raid rally car?

 

[Orion_SR]

Week end update.
9938, 100, 99, 99.
Just to concur with Orion_SR. Very frustating and my B spec driver still stuffs up.
Do I have to find 4 or 5 races where I overtook to quickly from a standing start?
What is the 2nd tune level for the RSC Raid rally car?

C-2: Toyota RSC Rally Raid '02, default + medium racing tires

62 missing points is too many for stray points from leading too quickly. I'm not all that sure how important that is for many tracks anyway. It is a prime concern for Nurburgring. At this point I'd lean toward missing an entire track, but you'd better finish all 8 tunings for floating points first.

Take all four cars classes, and maybe some alternate tunings, to Super Speedway to hunt for a few extra points. That track is stubborn and needs to be repeated even when no points are earned.