Diminishing returns will mean that the average $/g goes up as the build gets more and more extreme

I never understood the $/g metric. You want to minimize both, so having a high number is undesirable for cost, but desirable for weight. In my opinion, you should be calculating $ x g. That way you can minimize both. Of course, then you get into components that have the same $g value, but one is cheaper than the other but heavier, but that's exactly the kind of tradeoffs you're supposed to be evaluating!.
$/g only applies for REMOVING weight from a bike. As in, "I'm already 6.8kg and this $100 seatpost will drop 100g, but this $500 brakeset will drop 75g." Here you're trying to maximize g, while minimizing $, both which will lower your $/g ratio.

It's often being smart that can make a difference....

I was getting rid of an old kids bike and noticed the BB cable guide was single only, so took it off and weighed it, 4g saving over the conventional twin one I had had on my commuter, zero cost.

When I redid the brakes on my commuter I swapped the levers from UK to continental layout (front on left) as it suited a commuter (with hand signalling needs) better, the shortening this allowed on both cables (due to routing for rear) saved 19g at zero cost.

I swapped the levers for some bargain crosstop style ones I spotted (badly listed) on ebay, saving 60g and costing £2.20 ($3.50).

If you know the weight for every item on your bike, it's easy to spot the odd bargain 'save'.

_________________
Impoverished weight weenie wanna-be!
Budget 26" HT build viewtopic.php?f=10&t=110956

Some of you non-wrenching guys are not figuring the service cost in building the bike.
If you have built the bike yourself then you can arguably state the actual cost per gram.

But from here we can go to the maintenance cost per gram per year

_________________
I never took drugs to improve my performance at any time. I will be willing to stick my finger into a polygraph test if anyone with big media pull wants to take issue. If you buy a signed poster now it will not be tarnished later. --Graeme Obree

Re maintenance costs, what about replacing your aluminum and carbon cassette/chainring every 500 miles, and frequently snapping $1000 stem/seatposts?

Besides, the calculation is purely dependent on where you start anyway. I could achieve $0.1/g by moving from a 30lb bike ($100) into a 22lb bike ($500).

_________________
Fast falcons: http://www.youtube.com/watch?v=j3mTPEuFcWk" onclick="window.open(this.href);return false;
twitter: Falco_bike
Facebook: falco.bike

The calculation determines the endpoint. It's a marginal rather than average cost.

For example, suppose I have the following 3 options for a component, which are in every other way equivalent:

baseline: $0, 100 grams
A: +$50, 80 grams (-20 grams)
B: +$100, 70 grams (-30 grams)

You could argue B meets a $3.50/gram threshold, because it saves versus baseline 30 grams for $100, which is $3.33/gram. However, it does not, because versus A it saves 10 grams for $50, which is $5/gram. In the above, the $3.50/gram threshold is met only with A, which is $2.50/gram relative to baseline. The implicit A->B increment fails the test.

_________________
http://djconnel.blogspot.com/
Fuji SL/1
Send your data to Weight Weenie listings!

That's kind of what I was getting at. $/g for a whole bike makes absolutely no sense. It only applies for subtracting weight from a baseline, thus is really only useful if you already have a weight that you're trying to beat. It's only useful as a tool to compare upgrades on your current bike. So if your bike already weights 6.8kg and you have $3k to spend you can determine the most efficient use of your money to achieve the lowest possible weight. However, you cannot compare your $/g ratio to someone else lightening their bike unless they are also at exactly 6.8kg. Obviously the guy with the 8kg bike is going to have a much better $/g ratio when upgrading.

.... and if your baseline is 0 (as in, you're starting with frame, components, not upgrading from a previous existing build), then what? You're still saying that total $/g is irrelevant?

_________________
Exp001

Nice build, and a bargain at $0.68/ gr. Amazing weight for the dollars I think.
Well thought out build.
This was why I originally posted. Half knowing, there'd be gems out there.

_________________
2013 Giant TCR Advanced SL 0, 6.92kg
2013 Giant Defy Composite 2 M, 8.5kg - Wife's
Azzurro Torino 8.55g
Fuji 650 10.8kg
Miele Lupa Triple Tandem 38,89kg

Here's the problem:
spend $0 and drop 1000g, and your $/g goes up to $0.82. If we use the 6.8kg baseline, and calculate your "value ratio" as $3800/(6.8kg-5.671kg), your value ratio is $3.3658/g. Somebody else out there builds a 6.7kg bike for $380, their "value ratio" is $380/(6.8kg-6.7kg), or $3.8/g. It gets very skewed as you get closer to 6.8kg because it's a non-linear function, but at least reducing total weight moves the ratio in the right direction.

So what would be a better quantitative system then if the goal were to be lowest weight simultaneous to lowest total budget?
We've seen someone come up with an ultra-light, ultra-exclusive bike for plenty of money (gum's) - yet that just pushes what many believe to be the only way to 'go light', by throwing a lot of money at it. There must be some kind of quantitative value available to reinforce the notion that it is possible to go light, toe-to-toe with the majority of the "ultra light" and not be among the 1% that the 99% have been complaining about.

It really isn't all about "spend spend spend" but more "use your brain, ingenuity, intelligent and meticulous research, spend your money wisely".
So let's come up with a solid formula for figuring out an evaluation system based off of $ spent vs. lower weight.

_________________
Exp001

sorry to kill the thread!

Am I going to have to post pix of my receipts, to confirm? : )


C

Yes -- it's a basic economic principle. You balance marginal cost versus marginal benefit, not net cost versus net benefit.

Of course, given the choice between two bikes, each the same, you take the cheaper one, or at the same price take the lighter one, all other things equal. No marginal analysis required.

_________________
http://djconnel.blogspot.com/
Fuji SL/1
Send your data to Weight Weenie listings!

There can't be a truly definitive answer to this because there are two other factors to consider:

Rotating weight savings are more "valuable" that non-rotating weight. But by how much ?

Sometimes you can actually get a better $/gm ratio from a small component, but you would be willing to spend more $/gm because you can make a bigger overall weight reduction on a large component (such as a crankset) that you could simply never match with a series of smaller components.

So, in the end, there are personal preference and opinion factors.

I've got a whole spreadsheet for wheels, hubs, groupsets, contact points, saddles, handlebars, bottle cages, pedals..you name it.

From the numbers I've crunched, there are plenty of areas where you can upgrade on about 1 $/gram level. Once you take care of the 'low hanging fruit' i.e. pedals, groupset components, saddles, bottle cages I find that fancier stuff (e.g. carbon vs alloy handlebar, Campag vs SRAM) the ratio quickly rises to 3-4$ / gram.

Of course you also compromise on other aspects beside weight. Compared to a 'standard hub' (e.g. Dura-Ace or DT Swiss 240) the White Industries hub gives you much better $/gram than Alchemy. However the Alchemy flange dimensions are slightly different which potentially gives you a slightly stiffer wheel. Likewise when comparing a Velocity A23 to a HED C2 rim.

My new bike (under construction) will have titanium frame, SRAM Force groupset, Alchemy hubs laced to HED C2 rims, Time RXS pedals (titanium), King Ti bottle cages with alloy handlebars, stem and seatpost.

I blew my budget on the frame so I'll have 'cheap' contact points. Again that's an example of compromise. I save no weight with the Titanium frame however I spent more to get a custom build and (hopefully) more longevity and toughness.