Notices
RX-8 Discussion General discussion about the RX-8 that doesn't fit in one of the specialty forums.
Sponsored by:
Sponsored by: CARiD

Statistical Analysis Approach to Understanding MPG Issue

Thread Tools
 
Rate Thread
 
Old Oct 23, 2003 | 07:50 PM
  #26  
oosik's Avatar
Registered User
 
Joined: Oct 2003
Posts: 288
Likes: 0
From: Mobile, AL
1. 20
2. 10%
3. A
4. 113952
5. GT
6. 6spd
7. 93
8. yes
9. 440
10. 40%
11. 20%
12. FL
Reply
Old Oct 23, 2003 | 09:37 PM
  #27  
RodsterinFL's Avatar
Registered User
 
Joined: Jul 2003
Posts: 558
Likes: 0
From: Fort Myers, FL
1. 17.68 mpg
2. 0% Highway Driving (all town)
3. A
4. 101211
5. GT
6. 6 Speed
7. 87 octane
8. DSC/TCS = Fully Disabled
9. 3737 Miles
10. 90% Windows Open
11. 10% AC on (exact info this week)
12. Florida

Average speed per mile 29 mph!!!!
(added this from info at end of thread)

Last edited by RodsterinFL; Oct 24, 2003 at 10:00 PM.
Reply
Old Oct 23, 2003 | 09:44 PM
  #28  
mngpao's Avatar
Registered User
 
Joined: Oct 2003
Posts: 102
Likes: 0
From: North of Spokane WA
1. 22
2. 85%
3. A
4. 109104
5. GT
6. Automatic
7. 92
8. DSC/TCS = Yes
9. Miles to Date = 1650
10. Approximate % Time Driving with Windows Open = 0%
11. Approximate % Time Driving with AC on = 50%
12. State = WA

FYI with 95% hiway driving I get 24+ mpg.
Reply
Old Oct 23, 2003 | 10:11 PM
  #29  
Racer X-8's Avatar
Coming thru in waves...
 
Joined: Jun 2003
Posts: 1,488
Likes: 0
From: Somewhere between Yesterday and Tomorrow.
1. 21.0
2. 90%
3. B
4. 106429
5. GT
6. 6 speed
7. 93
8. Yes
9. 2267
10. 0%
11. 50%
12. SC

Kind of sounds like the Buckingham Pi theory...good luck...
Reply
Old Oct 23, 2003 | 11:22 PM
  #30  
sewingguy's Avatar
Registered User
 
Joined: Oct 2003
Posts: 7
Likes: 0
1. 21
2. 80
3. A
4. 112585
5. base
6. 6 speed
7. 93
8. No DSC
9. 2000
10. 95
11. 0
12. NY - feeling blue for the Yankees
Reply
Old Oct 23, 2003 | 11:34 PM
  #31  
Haze's Avatar
Registered
 
Joined: May 2003
Posts: 1,116
Likes: 0
From: Philadelphia, PA USA
Re: Statistical Analysis Approach to Understanding MPG Issue

1. RECENT Combined City/Highway Average MPG = 20
(Calculated Correctly and One Number please, not a range)

2. % Highway Driving = 15

3. Driving Style = A
A. Easy going (Don’t normally use high rpm range, only occasionally)
B. Average,
C. Above Avg (Use Upper RPM range quite often and accelerate hard)

4. VIN Number (last 6 digits) = 112741
(NOTE: If you already entered your VIN on the thread “Let's compare VIN numbers and fuel economy”, I can get it from that thread).

5. Model (eg. Base, GT, etc) = Base

6. Automatic / 6 speed = 6 Speed

7. Octane (Number) of Fuel Used = 93

8. DSC/TCS = No. Don't have it

9. Miles to Date = 1,571

10. Approximate % Time Driving with Windows Open = 25%

11. Approximate % Time Driving with AC on = 10%

12. State = PA
(I’ll later convert to Geographic Region)
Reply
Old Oct 23, 2003 | 11:59 PM
  #32  
8_wannabe's Avatar
Go baby!
 
Joined: Jun 2003
Posts: 1,303
Likes: 0
From: La Jolla CA
Norton -- Yer awesome. I nominate you for immediate instatement to Sr Member for at least attempting this. I am already impressed by your graph which shows a pretty good correlation btwn hiway driving and mpg. For those data points somewhat out of the norm, I bet if you query those drivers a bit more you would find out why they're so out of whack. For instance, "Highway driving" in LA is stop and go and could equate to city driving elsewhere. Recommend you investigate more those people who don't fall near the norm in that chart.

My responses:

1. RECENT Combined City/Highway Average MPG = 16.7
2. % Highway Driving = 40
3. Driving Style = B. Average to Above Avg (I accelerate hard but almost never go above 7K rpm. Cruise below 3500 rpm.)
4. VIN Number (last 6 digits) = It's on that thread.
5. Model: GT
6. 6 speed
7. Octane (Number) of Fuel Used = 87
8. DSC/TCS = Yes (fully enabled)
9. Miles to Date = 5100
10. Approximate % Time Driving with Windows Open: 50
11. Approximate % Time Driving with AC on: 40
12. State: CA
Reply
Old Oct 24, 2003 | 12:09 AM
  #33  
brothervoodoo's Avatar
RainMan is Back
 
Joined: May 2003
Posts: 2,650
Likes: 0
From: SF Bay Area
1. 15
2. 10%
3. C (Like someone mentioned before, I drive it like I stole it).
4. 102787
5. GT
6. 6 speed
7. 91
8. DSC/TCS = off
9. Miles to Date = ~5,500
10. 30%
11. 1%
12. State = CA
Reply
Old Oct 24, 2003 | 12:10 AM
  #34  
ptiemann's Avatar
Registered User
 
Joined: Sep 2003
Posts: 493
Likes: 0
From: San Francisco Bay Area
For instance, "Highway driving" in LA is stop and go and could equate to city driving elsewhere.
I think stop-and-go is actually worse than city driving. When I put 95% highway I was thinking of the tanks where I avoided commute hours. Silicon Valley traffic is not much different from LA.
Reply
Old Oct 24, 2003 | 03:08 AM
  #35  
druck's Avatar
UK Owners Club
 
Joined: Sep 2003
Posts: 402
Likes: 0
From: Gloucester, England
Originally posted by ptiemann
I think stop-and-go is actually worse than city driving. When I put 95% highway I was thinking of the tanks where I avoided commute hours. Silicon Valley traffic is not much different from LA.
Stop and go is bad, not only do you need more fuel to pull off each time, but also sitting stationary uses as much fuel as traveling at around 12mph, and you obvious aren't putting any miles on the clock.

Cheers
---Dave
Reply
Old Oct 24, 2003 | 07:29 AM
  #36  
8_wannabe's Avatar
Go baby!
 
Joined: Jun 2003
Posts: 1,303
Likes: 0
From: La Jolla CA
Originally posted by druck
Stop and go is bad, not only do you need more fuel to pull off each time, but also sitting stationary uses as much fuel as traveling at around 12mph, and you obvious aren't putting any miles on the clock.
Exactly my point. If you say your are driving 50 percent hiway but your hiway is stop and go, might as well count it like city driving. It is assumed hiway driving is more efficient (i.e, greater miles per gallon) but clearly that is not always the case.
Reply
Old Oct 24, 2003 | 10:27 AM
  #37  
norton's Avatar
Thread Starter
Registered User
 
Joined: Aug 2003
Posts: 42
Likes: 0
From: Central Jersey
For instance, "Highway driving" in LA is stop and go and could equate to city driving elsewhere. Recommend you investigate more those people who don't fall near the norm in that chart.
Exactly my point. If you say your are driving 50 percent hiway but your hiway is stop and go, might as well count it like city driving. It is assumed hiway driving is more efficient (i.e, greater miles per gallon) but clearly that is not always the case.
On my 42 mile Central Jersey commute to work this morning I did a lot of pondering about this % Hwy thing. At the outset, it appears like it may be a pretty important factor impacting MPG. There are a couple shortcomings in using % Hwy numbers though. These are described as follows:

1. Per the quotes above (great observation), not all Hwy driving is the same type of driving for everyone and its not the same for an individual every time that person hits the highway. Likewise not all city driving is the same.

2. The % Hwy is an educated guess, at best a ballpark figure. It's not really an emperically measured number like odometer reading or mpg. There is bound to be some error in these estimates.

3. People may have differing interpretations as to exactly how Highway and City are defined. Thus a somewhat apples to oranges comparison among responses.

Even with these problems though, I am fairly pleased with the results so far, using % Hwy as one of the primary determining factors of MPG.

*************************************************

I do have a NEW variable I thought of, which pretty much removes the shortcomings of using % Hwy described above. Unfortunately determining this factor will be a pain, time consuming, and will take a while before you could provide it. The NEW factor is............. . Average Speed per Mile .

The good thing about using Avg Speed per Mile is that it is totally emperical. No subjective guessing involved, no different interpretations of Hwy vs City, and it pretty much "normalizes" all road conditions (eg A highway full of stop and go traffic is quite similar to city driving).

The pain in using Average Speed per Mile is in the calculation, though I will simplify it as much as possible. Basically you would need to provide a couple things:

[/B]1. # of miles traveled recently (from odometer reading).

2. The amount of time spent driving. (unfortunately this requires recording the beginning and ending time of every ride).[/B]

Average Speed Per Mile = (Total Miles / Total Minutes) * 60

This calculation should be done on the same odometer mileage as is used in calculating MPG. (Yes, unfortunately that would need to be provided again).

I'm NOT suggesting we do this right now (or even that we do it), however it is certainly something to think about. My guess is that there would be a very very strong relationship between Avg Speed/Mile and MPG.
Please let me know what your thoughts and feelings are about later pursuing the Avg Speed/Mile information. If people think its overkill or not worth it, I can understand their viewpoint.

*******************************************

For now I would like to stay focused on working with the great data I have been getting. I have begun working on the Regression Model. Due to the volatility and variation in the observations I'm receiving, the Regression results vary somewhat with each additional response. As I continue getting more observations, the results should begin to settle towards a stable conclusion.

Great response data. Please keep 'em coming. Thanks.
Reply
Old Oct 24, 2003 | 10:36 AM
  #38  
Racer X-8's Avatar
Coming thru in waves...
 
Joined: Jun 2003
Posts: 1,488
Likes: 0
From: Somewhere between Yesterday and Tomorrow.
Too bad we don't have an hour meter for total ignition-on time. We could then just use the odometer for miles & total gas consumed (if you've been recording that, like you should if you're really recording things to compute overall mpg).
Reply
Old Oct 24, 2003 | 04:02 PM
  #39  
red_rx8_red_int's Avatar
Registered User
 
Joined: Jul 2003
Posts: 911
Likes: 1
From: NC
1. 20.2
2. % Highway Driving = 90
3. Driving Style = C
4. VIN Number (last 6 digits) = 103169
5. Model (eg. Base, GT, etc) = GT
6. Automatic / 6 speed = 6sp
7. Octane (Number) of Fuel Used = 93
8. DSC/TCS = Yes
9. Miles to Date = 4700
10. Approximate % Time Driving with Windows Open =0
11. Approximate % Time Driving with AC on = 50
12. State = MO
Reply
Old Oct 24, 2003 | 06:06 PM
  #40  
vibec's Avatar
Registered User
 
Joined: Sep 2003
Posts: 27
Likes: 0
From: Perth, Australia
1. RECENT Combined City/Highway Average MPG = 15
(Calculated Correctly and One Number please, not a range)

2. % Highway Driving = 20%

3. Driving Style = C
A. Easy going (Don’t normally use high rpm range, only occasionally)
B. Average,
C. Above Avg (Use Upper RPM range quite often and accelerate hard)

4. VIN Number (last 6 digits) = 574
(NOTE: If you already entered your VIN on the thread “Let's compare VIN numbers and fuel economy”, I can get it from that thread).

5. Model (eg. Base, GT, etc) = GT

6. Automatic / 6 speed = 6MT
7. Octane (Number) of Fuel Used = 95

8. DSC/TCS = Yes / No(Don’t Have or Fully Disabled)yes

9. Miles to Date = 800

10. Approximate % Time Driving with Windows Open = 10%___________

11. Approximate % Time Driving with AC on = _30%

12. State = wa
(I’ll later convert to Geographic Region)
Reply
Old Oct 24, 2003 | 06:43 PM
  #41  
rotarynews.com's Avatar
Registered
 
Joined: Aug 2002
Posts: 274
Likes: 0
From: Viva Las Vegas!
1. 16.5mpg avg last 4 fill ups (with an autox)
2. 40%
3. C ++
4. 101182
5. Touring
6. 6 speed
7. 91
8. dsc off 50% time
9. 4400
10. 20%
11. 80%
12. NV
Reply
Old Oct 24, 2003 | 08:37 PM
  #42  
norton's Avatar
Thread Starter
Registered User
 
Joined: Aug 2003
Posts: 42
Likes: 0
From: Central Jersey
Well I have 24 observations so far. That's about the minimum I need, so I will hopefully get the initial Regression Analysis done tomorrow. Here is an updated graph depicting % Hwy vs MPG. Still shows a definite pattern.

One interesting thing to note are the two outliers which show 20 mpg, yet a low % Hwy - 10% and 15% respectively. Based on % Hwy alone, someone would probably conclude those are aberations. There is a valid reason for these two outliers though. For those two observations, the Driving Style was "A" (easy going), thus yielding a higher MPG. This is exactly the kind of situation that the Regression Analysis automatically takes into account, by simultaneously assessing multiple factors.

Can't wait to get the Regression results.

Reply
Old Oct 24, 2003 | 08:46 PM
  #43  
lurcher's Avatar
ex-preorderer
 
Joined: Mar 2003
Posts: 171
Likes: 0
From: Dundee, Scotland
Erm, hate to throw a spanner in the works at this stage, but is that % highway driving by time or by distance?
Reply
Old Oct 24, 2003 | 09:26 PM
  #44  
norton's Avatar
Thread Starter
Registered User
 
Joined: Aug 2003
Posts: 42
Likes: 0
From: Central Jersey
Erm, hate to throw a spanner in the works at this stage, but is that % highway driving by time or by distance?
Lurcher: Excellent catch. You've got a very valid point. Let's see now.

If someone spent only 5% of their time on the highway and drove 80% of the total combined hwy/city distance on th hwy, would we say that person has 5% hwy proportion or an 80% hwy proportion? This situation would mean that the person spent a frustrating 95% of their driving time going very little distance (20% of the combined hwy/city distance).

My view is that the % Hwy would be based on DISTANCE. The reason is that City Driving implicitly assumes more time spent going fewer miles than Hwy Driving. So I believe the time factor in the % split is already taken into consideration.

Nonetheless, you've raise a very good point about the ambiquiety in using % Hwy. I think it's an o.k. variable to use in the analysis, however Average Speed per Mile is an infinitely better measure to assess City vs Hwy driving. (Please see my comments about 7 posts up.
Reply
Old Oct 24, 2003 | 09:35 PM
  #45  
8_wannabe's Avatar
Go baby!
 
Joined: Jun 2003
Posts: 1,303
Likes: 0
From: La Jolla CA
Keep rockin', Norton. You data and methodology seem very valid so far.
Reply
Old Oct 24, 2003 | 09:46 PM
  #46  
norton's Avatar
Thread Starter
Registered User
 
Joined: Aug 2003
Posts: 42
Likes: 0
From: Central Jersey
Thanks 8_wannabe. I've been following this terrific forum for close to a year always find it interesting and exciting. I would really like to help towards understanding the factors heavily impacting mpg.
Reply
Old Oct 24, 2003 | 09:48 PM
  #47  
red_rx8_red_int's Avatar
Registered User
 
Joined: Jul 2003
Posts: 911
Likes: 1
From: NC
Originally posted by norton
Lurcher: Excellent catch. You've got a very valid point. Let's see now.

If someone spent only 5% of their time on the highway and drove 80% of the total combined hwy/city distance on th hwy, would we say that person has 5% hwy proportion or an 80% hwy proportion? This situation would mean that the person spent a frustrating 95% of their driving time going very little distance (20% of the combined hwy/city distance).

My view is that the % Hwy would be based on DISTANCE. The reason is that City Driving implicitly assumes more time spent going fewer miles than Hwy Driving. So I believe the time factor in the % split is already taken into consideration.

Nonetheless, you've raise a very good point about the ambiquiety in using % Hwy. I think it's an o.k. variable to use in the analysis, however Average Speed per Mile is an infinitely better measure to assess City vs Hwy driving. (Please see my comments about 7 posts up.
I think it's OK as it is as long as people correct for true highway vs city highway driving. I think most people would classify stop and go rush hour traffic on a highway/freeway as city driving. If this is so then this is a non-issue. But another issue is that not all highway driving is equal. I logged a lot of highway miles today with an average speed of probably about 83, which should result in a much lower MPG than if I had gone 55. So maybe we should add an average highway speed to the matrix. Note 25 mph and 35 mph won't really matter, so we do not need an average city speed.
Reply
Old Oct 24, 2003 | 10:05 PM
  #48  
RX-GR8's Avatar
Registered
 
Joined: Aug 2003
Posts: 3,098
Likes: 3
From: Cherry Hill, NJ
1. RECENT Combined City/Highway Average MPG = _15.8__________
(Calculated Correctly and One Number please, not a range)

2. % Highway Driving = __10%____

3. Driving Style = B
A. Easy going (Don’t normally use high rpm range, only occasionally)
B. Average,
C. Above Avg (Use Upper RPM range quite often and accelerate hard)

4. VIN Number (last 6 digits) = ___________
(NOTE: If you already entered your VIN on the thread “Let's compare VIN numbers and fuel economy”, I can get it from that thread).

5. Model (eg. Base, GT, etc) = ____GT_______

6. Automatic / 6 speed = ____6_________

7. Octane (Number) of Fuel Used = ___93_________

8. DSC/TCS = Yes / No(Don’t Have or Fully Disabled) fully disabled when i remember

9. Miles to Date = ___1018________

10. Approximate % Time Driving with Windows Open = ____50_______

11. Approximate % Time Driving with AC on = ______5%_____

12. State = ___NJ___________
(I’ll later convert to Geographic Region)
Reply
Old Oct 25, 2003 | 02:18 AM
  #49  
canzoomer's Avatar
Forum Vendor
 
Joined: Jul 2003
Posts: 1,223
Likes: 1
From: Edmonton, Alberta
1. RECENT Combined City/Highway Average MPG = 15.044
2. % Highway Driving =65%
3. Driving Style = C
4. VIN Number (last 6 digits) =101877
5. Model (eg. Base, GT, etc) = GT
6. Automatic / 6 speed = 6 speed
7. Octane (Number) of Fuel Used = 91
8. DSC/TCS = Yes
9. Miles to Date = 3,240M/5217km
10. Approximate % Time Driving with Windows Open = 2
11. Approximate % Time Driving with AC on = 0
12. State = Alberta, Canada

Can I mention that it would be more useful if you dealt with kilometers, liters and other than only US geography?

I also suggest to ask altitude, as it has a significant effect on fuel economy.
In my case 2,700ft/855M
Reply
Old Oct 25, 2003 | 02:19 AM
  #50  
canzoomer's Avatar
Forum Vendor
 
Joined: Jul 2003
Posts: 1,223
Likes: 1
From: Edmonton, Alberta
Originally posted by matt
Here's another take on the issue. What if the speedo's are.."optomistic"? say 1.5 mph off. That would reduce the MPG. I think before you can establish a baseline. You need to use another method of calculation. If the speedo's are at fault, then all the data collection in the world will not show a thing.


my .02 yen
Speedos are 4% optimistic on the stock 18" wheels/tires.
Reply


You have already rated this thread Rating: Thread Rating: 0 votes,  average.


All times are GMT -5. The time now is 05:32 PM.