Thank you Jerry.  I was quoting the scale the manufacturer quotes.  If you divide 15mm into one foot, or 304mm, you get 1:20.2666 recurring.
Either way, Bachmann Spectrum, at 15mm to the foot on 45 mm gauge track is probably as close to accurate as we are going to see in the large scales.

The point about HO versus OO is relevant, but it goes back to the difficulty of making small motors back when the scale/gauge was struck.  The point I made about 1:48 O scale was not that the gauge was exact, but that O scale narrow gauge uses narrower track, like the prototype, not larger rolling stock on the same track.

My original intention, which has been lost, was to give the original poster some information with which to start making his purchases.  If he wants an eclectic collection of shiny toys in the garden, that is his prerogative.   But if, like me, he wants to start a collection of appropriately sized models, which he can use to follow a prototype with some degree of accuracy, he now has some useful information.  Many newcomers to the hobby buy items they later regret purchasing when their knowledge base grows and they become more discerning.  I hoped to help him, or others, avoid that.

Current day modelers are in a position where we should not be limited by technology or knowledge.  There are publications encouraging people to become Master Model Railroaders and spread their experience, techniques and wisdom.  Why am I a pariah for pointing out things which can lead to more accurate modelling?

The original poster asked whether G scale was standard gauge or narrow gauge.  Clearly, if referring to 1:29 'scale' the answer must be both.

1:29 models try to bridge a gap between standard and narrow gauge equipment.  This is a gap which should never be bridged.   If I want toy trains I would play with Brio or Lionel.  I thought the poster wanted some accurate information to actually model a true scale to gauge ratio, that is what I tried to offer.

I was looking for a forum for scale model railways.  I must be in the wrong place.

Armorsmith, You have hit on the single biggest problem with large scale model railways.  It is commonly known as 'G' scale, but that is a misnomer.

Every other scale uses a scale ratio and then sets the track gauge to differentiate standard or narrow gauge.

For example, 'O' scale (in the USA) is 1:48 (or one quarter inch scale)  (Ignore Lionel etc, I am talking two rail, accurate scale models)  If you run equipment on 32mm gauge track you have standard gauge.  If you run On30, you still measure a quarter inch to the foot for building, but the track is 16.5mm.  This equates to 30 inch or 2 foot, six inch.  On3 uses 19mm (i think - whatever the 3/4 inch equivalent is) and it accurately reflects three foot narrow gauge.  For most people this makes sense.

Garden railways typically use 45mm track for all large scales.  (The English use 32mm as well to represent narrower gauges)  The first mass manufacturer in 45mm track was LGB, and they chose to model one meter narrow gauge (39.37 inches) so the scale ratio is 1:22.5.

Early American narrow gauge modelers took the 45mm track and modeled in 1:24.  This is a great scale for building rolling stock or structures, as it is one half inch to the foot.  (Easy conversion.)  However it equates to 42 inch or 3 foot 6 inch.  Not a popular gauge in the States.

For 45mm track, standard gauge equipment would be 1:32, or 10mm equals one foot.  Popular in England in the early days of model railways and easy for building, using a metric ruler.

Bachmann began modelling narrow gauge in 1:24, but improved accuracy with the Spectrum range, which is 1:20.3, or 15mm to the foot.  This accurately reflects three foot gauge on 45mm.  [Bachmann took a lead here and should be congratulated for it.  Accuracy of scale/gauge is lacking elsewhere.]

The different scale/gauge combinations are set out below.  Measurements are relative to 45mm track gauge.

1:32 = 56.69 inches or 4 foot 8 and 5/8 inch.  (Just 1/8th inch over.)  I consider this too close for argument.

1:24 = 42.52 inches or 3 foot 6 and 1/2 inch (1/2 inch over)  Again, close enough.

1:22.5 = 39.86 inches or 1.0125 metres.  (12.5mm over, just under 1/2 inch) No argument.

1:20.3 = 35.965 or 1/32nd less than three feet.  Almost perfect.

Now we strike the problem!

1:29 = 51.378 inches or 4 foot 3 and 3/8 inches.  What railroad ran on 4 foot 3 and 3/8th?  It is basically 10% smaller than standard gauge 1:32
and 20 % bigger than 1:24.  It is a nothing scale/gauge combination.

If trying to model a particular railroad, you should shop for items which list a scale/gauge ratio which matches your prototype.  Quality manufacturers advertise this.  You will find that anything listed as 'G' scale is probably 1:29 or has no accurate prototype and is simply built to look similar to currently manufactured equipment.  I refuse to support these manufacturers.

Sorry this is so long.  Hope it helps.

Wow! I really opened a can of worms with this.  I will probably purchase online.  The information provided should help me get what I am looking for.  Thank You.

My logging layout runs on  track powered, 12 volt DC.  When working properly, everything runs slow and looks the way it should.  Most of my locos are kit-bashed or scratch-built using LGB mechanisms or short wheelbase, four wheel power blocks from Aristo. Everything runs slow and smooth, provided the track is lightly cleaned and there is sufficient weight in the loco. 

Everything was fine until Bachmann started making geared locos, and I had to have some.

I bought an original two truck Shay, which required new die-cast trucks, but is still going strong. 
I then bought an original Climax which required re-wiring between each truck and the frame to get it to run reliably. 
And, although I should have learned, I bought an original version 2-4-2 Southern.  I ended up putting a six wheel drive mechanism from an LGB 0-6-2T loco in it to make it run slowly and smoothly.

I also found an original ten-wheeler cheap. Very cheap.  I now know why.  The mechanism in this is a complete waste of space, and will one day be replaced or kitbashed.

When I saw the Porter reviewed recently I figured it was safe to buy one...but make sure it is a recent one!

Thanks again,

Mick

I am interested in purchasing a Large Scale Bachmann Porter.  The current catalogue number is 82598. 

Was this the number for the original release?  How do I tell if an advertised model is the new or old release.

I want the new, better running mechanism.

Mick