Pointless Statistics

[Alex Bridgeman]

As it's coming close to the end of the year, it seems to be traditional to take a look at any statistics that I can come up with, which can be vaguely related to port.

There is, of course, the classic statistic of time taken by the cellar to gain 1 collective year of additional bottle age, but following the debates of last year I now accept that this is exactly 12 months.

Given my significant withdrawls from the cellar and modest acquisitions this year, the most interesting statistics (to me, at least) are about what I have left in the cellar. 82% of my bottles are port, 7% is bordeaux, 2% Australia, 2% English and the rest is somewhat of a mix.

The 21st anniversary of any particular declaration is traditionally considered the time it is ready for drinking. On this basis, 42% of my port is ready and the rest is still to be left to slumber. Of this 42%, 13 comes from the 1963 vintage (no surprise, really). 1994 accounts for 9% of my port and 1997 accounts for 8%. A surprising (at least it is to me) 20% of my port comes from the 21st century.

My second-favourite shipper is Fonseca, and they make up 10% of my cellar. Graham, Smith Woodhouse and Martinez each account for 5%. It will come as no surprise to anyone that Quinta do Vesuvio is the largest shipper in my cellar - but guess what the percentage turns out to be...

[jdaw1]

The first Vesuvio vintage was 1989, which isn’t yet ready, so all the V lies in the 58% (of the port) slumber zone. We also know that Fonseca is 10% of the port (let me guess that is mostly vintages up to 1985, so not reducing the 58%), and 5% each for Graham, Smith Woodhouse and Martinez. I know that you have ≥2 bottles of Mz97, and some SW90s and 2k’s also plausible. So let’s say that half of that 15% isn’t yet mature, leaving about 50% which could be Vesuvio. Leave off some Mg91 (lots?), leaves a Vesuvio percentage between 10 and 40.

So, having thought too hard I’ll just guess at 25%.

[ajfeather]

Vesvuio is also the most popular port in my cellar (pipping Niepoort). Given I have nearly twice as much vesuvio as fonseca I will go for 20%!

[Alex Bridgeman]

With a mixture of shame and pride, I can tell you that you are both too low - but that Julian's logic was working up until the point that he threw in the towel and guessed.

[SushiNorth]

Hmm, i never bothered to do my VP math... but you've inspired me. Excluding some pending purchases...
@ 12% Warres, Graham
@ 7-9% Taylor, Fonseca, Calem, Croft
@ 1-6% 15 other houses.
By the 21yr measure, only 8% is drinkable and all of it is SW, Grahams, and Warres.

[Alex Bridgeman]

I admit, all these statistics are reported by the spreadsheet I developed to track the contents of my cellar. I find it mildly amusing to be able to find out what proportion of my cellar is port. How much of it is ready to drink.

I wonder what proportion of my bottles become drinkable on 1 January under the 21 year rule...

[SushiNorth]

I wonder what proportion of my bottles become drinkable on 1 January under the 21 year rule...

88 was a decent year... I pick up another 4%, but I've been drinking those this past year already.

[uncle tom]

Currently 2,485 bottles of VP, with average age of 25.684 years, total bottle years 63,825, increasing by one year every three hours and thirty two minutes..

At present, I have stocks of 285 different vintage ports.

..I might concede the word 'adequate' at some point, but not just yet...

Tom

[jdaw1]

with average age of 25.684 years

Are you willing to report deciles?

increasing by one year every three hours and thirty two minutes..

! by one bottle-year every !; and anyway merely being one year divided by the number of bottles.

[g-man]

hmmm

18% of my collection is port
out of which 90% of my collection is Fonseca.
5% taylors
5% dow

[Glenn E.]

Given JDAW's math and AHB's reputation, I will venture a guess at 33%.

[uncle tom]

Are you willing to report deciles?

That doesn't work terribly well - for example, the oldest 10% of bottles includes slightly less than half of my '63's - which bottles does one include?

Tom

[Glenn E.]

I created a spreadsheet from memory... it should be pretty close, but is certainly off by a handful of bottles. When one's "cellar" is in fact a refrigerator that in the best circumstances will hold roughly 200 bottles, something as powerful as CellarTracker is overkill.

45% of my cellar is too young to drink.
65% of my cellar is VP or SQVP.

An overwhelming 88% of my cellar is 750 ml bottles. I have but two 500ml bottles and five Magnums, but nearly 8% of my cellar is half bottles.

17% of my cellar is Quinta do Noval, which is apparently my favorite producer. Graham is second at 13%, followed by Rocha (12%), Dow (10%), Vesuvio (8%), and Taylor (7%). (I'm double counting Magnums for this purpose, but still counting halves as bottles. Rocha drops just below Dow if those 8 halves are counted as 4 bottles.)

1994 is apparently my favorite vintage at 20%, this despite the fact that I have never opened a bottle of 1994. 1964 follows closely at 19%, then 2003 (13%) and 1985 (12%) and 1977 (10%).

Edit: updated spreadsheet with actual contents of the fridge, so it's accurate now I didn't count non-cellared bottles such as the ruby reserves and 20-yr old tawnies that I have stored in a handier location.

Edit2: 2008 - 1977 != 21. Even more of my cellar is of drinking age than I previously thought! Though personally I do think that 30 years is a more appropriate limit than 21 years, so for me it's still 61% of my cellar that is too young to drink.

[jdaw1]

Are you willing to report deciles?

That doesn't work terribly well - for example, the oldest 10% of bottles includes slightly less than half of my '63's - which bottles does one include?

Just the deciles of the years: so one would report ‟1963”.

(Or, pedantically, one could order the houses by the date that their harvests finished too much like work.)

[uncle tom]

65% of my cellar is VP or SQVP.

- 95.3%

An overwhelming 88% of my cellar is 750 ml bottles. I have but two 500ml bottles and five Magnums, but nearly 8% of my cellar is half bottles.

- 99.6% are 750ml - I have just six magnums, one litre and two halves

(but after tomorrow, there will be just one half!)

Tom

[Glenn E.]

65% of my cellar is VP or SQVP.

- 95.3%

Yes, but this still means you have more than twice as many bottles of non-VP as I have.

4.7% of your cellar is 117 bottles. 35% of my cellar is... 51 bottles.

[uncle tom]

Yes, but this still means you have more than twice as many bottles of non-VP as I have.

Yes, I have a fairly varied mix of non-vintage ports, including an interesting collection of Crusteds going back to 1941, and some interesting tawnies, two of which pre-date the 'indication of age' era. I also have a modest stash of colheitas and an adequate stock of LBV's. I have only four bottles of 'standard' port, and three of those are whites..

Tom

[Alex Bridgeman]

Given JDAW's math and AHB's reputation, I will venture a guess at 33%.

Almost spot on - it's 34%! That's a lot of Vesuvio - what do I do with all those pretty little chestnut boxes?

[Alex Bridgeman]

Currently 2,485 bottles of VP, with average age of 25.684 years, total bottle years 63,825, increasing by one year every three hours and thirty two minutes..

At present, I have stocks of 285 different vintage ports.

..I might concede the word 'adequate' at some point, but not just yet...

Tom

The most important statistic that I keep in my mind is the one where I calculate the number of bottles I will need to buy in the future to have sufficient port to drink at my current rate for the rest of my life (assuming that I live to the average age of my father and both grandfathers). Frighteningly, I still need more bottles than Tom has in his cellar...

[Alex Bridgeman]

I wonder what proportion of my bottles become drinkable on 1 January under the 21 year rule...

88 was a decent year... i pick up another 4%, but I've been drinking those this past year already heheh.

I've worked it out. I only gain another 2% of my cellar for current drinking

[Alex Bridgeman]

I've played around with my spreadsheet a little more and come up with some more statistics.

96% of my port is vintage or SQVP, 2% is crusted and 2% is unfiltered LBV. I have one bottle each of colheita, ruby and tawny.

96% of my port containers are 750ml bottles. All of the remaining 4% are half bottles. I own no magnums (and can't really see this changing considering the way in which I store and drink port).

And on the "thirty-years for drinking rule", only 35% of my port would be ready for drinking. At my normal rate of consumption, that will last me just under 3 years. Yikes! I need to go shopping after Christmas!

[Glenn E.]

I consume at a much slower rate than the rest of you. I probably average one bottle of VP per month over the course of a year, plus another bottle of Colheita every other month. So at total of 18 bottles out of my cellar in any given year. (And really, that's probably too high at my current rate of consumption. But I'm allowing for an assumed rate increase. )

The bulk of my consumption is Tawnies with an indication of age, which I purchase and drink right away. I don't have any of those included in my cellar spreadsheet because they wouldn't stay there long enough to matter.

So... just counting VP and Colheita... my cellar will last about 8 years. Eek! I'm almost out of Port!!!

[Alex Bridgeman]

Whilst watching something pointless on TV this evening, I was playing with my tasting note database and decided to pull out some pointless statistics. Pointless, because I will not do anything as a result of having worked out the spread of tasting notes written over the course of the 12 months of 2010. I found that of the wines I scored, the distribution was:

>97 points - not used during 2010
97 points - 1 wine (0.3%)
96 points - 1 wine
95 points - 6 wines (1.7%)
94 points - 12 wines (3.4%)
93 points - 14 wines (4.0%)
92 points - 32 wines (9.1%)
91 points - 34 wines (9.7%)
90 points - 33 wines (9.4%)
89 points - 41 wines (11.7%)
88 points - 34 wines (9.7%)
87 points - 25 wines (7.1%)
86 points - 19 wines (5.4%)
85 points - 33 wines (9.4%)
84 points - 19 wines (5.4%)
83 points - 12 wines (3.4%)
82 points - 6 wines (1.7%)
81 points - 5 wines (1.4%)
80 points - 8 wines (2.3%)
79 points - 3 wines (0.9%)
78 points - 2 wines (0.6%)
76 points - 1 wine (0.3%)
72 points - 1 wine (0.3%)

Does this seem a reasonable distribution of scores?

[DRT]

Does this seem a reasonable distribution of scores?

Yes, apart from a small blip at 86 points your scores appear to more or less follow a classic standard disctibution curve.

[JacobH]

I decided to pull out the same pointless statistics as AHB. I rate Ports by my answer to the to the question ‟Is this a good Port?”. There are three categories Certainly, Possibly, and No. When I am feeling particularly equivocal I might add a plus or minus sign. The results are:

C+ 27
C 83
C- 4

P+ 15
P 77
P- 16

N+ 0
N 19
N- 2

Comparing my statistics to AHB’s (assuming that 90+ points ≈ C, 83+ ≈ P, <=82 ≈ N), it appears I have a somewhat less critical palate than he does, since I certainly don’t think I drink better Port than he!

[Alex Bridgeman]

I've played around with my spreadsheet a little more and come up with some more statistics.

96% of my port is vintage or SQVP, 2% is crusted and 2% is unfiltered LBV. I have one bottle each of colheita, ruby and tawny.

96% of my port containers are 750ml bottles. All of the remaining 4% are half bottles. I own no magnums (and can't really see this changing considering the way in which I store and drink port).

And on the "thirty-years for drinking rule", only 35% of my port would be ready for drinking. At my normal rate of consumption, that will last me just under 3 years. Yikes! I need to go shopping after Christmas!

Oh how things change - how naive I was when I wrote this.

Today, 2 years later, things are different and I have benefited from having been shopping. I am now down to only 93% of my port being vintage or SQVP thanks to the ludicrously cheap offers of Croft LBV that Tesco have made at Christmas in each of the last two years. This has boosted my LBV stocks to account for 5% of my port, Crusted is 2% and I have 2 bottles of Tawny with an indication of age, 3 bottles of Tawny with no indication of age, one bottle of white port, two bottles of ruby port and three bottles of colheita. All together, there are 358 different ports represented in my cellar. 88% of the corks in my cellar are in bottles, 9% are in half bottles, 3% are in magnums and there are 3 screw-capped bottles. 40% of my port is aged 30 years or more and I nearly have enough to keep me going for the rest of my days provided I don't increase the rate at which I drink! I just need to buy a few bottles each new release in case I have American visitors drop by to see me.

[Alex Bridgeman]

Almost a year on from my last post, I've decided it's time to post some updated statistics - this time based on my tasting notes.

In 2005 I tasted 45 different ports
In 2006 I tasted 153 different ports
In 2007 I tasted 157 different ports
In 2008 I tasted 278 different ports
In 2009 I tasted 339 different ports
In 2010 I tasted 351 different ports
In 2011 I tasted 470 different ports

In 2006 the average age of port I tasted was 19 years old
In 2007 the average age of port I tasted was 39 years old
In 2008 the average age of port I tasted was 36 years old
In 2009 the average age of port I tasted was 29 years old
In 2010 the average age of port I tasted was 29 years old
In 2011 the average age of port I tasted was 30 years old

And the distribution of my scores in 2011 was:
100 - 0
99 - 2
98 - 3
97 - 1
96 - 3
95 - 7
94 - 16
93 - 26
92 - 36
91 - 34
90 - 40
89 - 47
88 - 42
87 - 36
86 - 37
85 - 28
84 - 20
83 - 15
82 - 12
81 - 4
80 - 8
79 - 3
78 - 4
77 - 1
76 - 1
75 - 0
74 - 2

Anybody interested in updated statistics from my cellar spreadsheet?

[DRT]

Anybody interested in updated statistics from my cellar spreadsheet?

Yes please.

[Alex Bridgeman]

So, some headline (and pointless) statistics from my cellar spreadsheet:

91% of my bottles (as in "glass container of any size used to contain wine of alcoholic content 3-20%") contain Port, 3% are claret, 2% are Champagne, 1% are Australian, 1% are English and the balance are from Spain, Hungary, South Africa, other parts of France, and other parts of the world. I have 4 bottles of Madeira.

Of the bottles with an indication of age, the average age of the wine in my cellar is 31 years and 114 days.

Of the Port, 89% are bottles, 8% are half bottles and 3% are magnums. There is nothing larger than a magnum.

There are 399 unique Ports in my cellar.

69% of my Port is vintage, 25% is single quinta vintage (which includes Vesuvio), 4% is LBV (mostly Croft 2004!) and 2% is Crusted.

My top 3 shippers are Vesuvio (16%), Fonseca (7%) and Graham (7%).

My top 3 vintages are 1963 (10%), 1994 (7%) and 1977 (6%).

And finally, using the 21 year rule for judging port to be ready for drinking, next year 59% of my port cellar will be ready

[RAYC]

Almost a year on from my last post, I've decided it's time to post some updated statistics - this time based on my tasting notes.

In 2005 I tasted 45 different ports
In 2006 I tasted 153 different ports
In 2007 I tasted 157 different ports
In 2008 I tasted 278 different ports
In 2009 I tasted 339 different ports
In 2010 I tasted 351 different ports
In 2011 I tasted 470 different ports

This is the most impressive statistic! 9 per week on average...up from below 1 per week in 2005...

[Alex Bridgeman]

Almost a year on from my last post, I've decided it's time to post some updated statistics - this time based on my tasting notes.

In 2005 I tasted 45 different ports
In 2006 I tasted 153 different ports
In 2007 I tasted 157 different ports
In 2008 I tasted 278 different ports
In 2009 I tasted 339 different ports
In 2010 I tasted 351 different ports
In 2011 I tasted 470 different ports

This is the most impressive statistic! 9 per week on average...up from below 1 per week in 2005...

And now it's up to 473 - 3 more ports tasted so far this week.

[Alex Bridgeman]

I was bored last night and added some functionality to my tasting notes statistics spreadsheet. I can now tell you that since I started keeping my notes electronically in 2005, I have tasted 1902 bottles of port (including last night's Churchill '82).

The most frequently tasted shippers come as no surprise, and are:
Taylor (inc. Vargellas and Terra Feita) 8%
Warre (inc. Cavadinha and LBV) 8%
Graham (inc. Malvedos) 8%
Fonseca (inc. Guimaraens) 7%
Noval (inc. Silva and Nacional) 6%
Dow (inc. Bomfim and Riberia) 6%
Croft (inc. Roeda and LBV) 5%
Vesuvio (inc. Capela) 5%

These 8 producers account for over half of all the port that I drink! It is also a pretty representative list of the ports that I will seek out and buy when I'm looking to add to my cellar. I like mature Cockburn and Sandeman vintage ports, but these are not that easy to find in the UK at the moment.

[benread]

Vesuvio (inc. Capela) 5%

Given your well known preference for mature vintage port, and the relative youth of vesuvio (unless you have a large stash of the 1890'something bottles) I was surprised it was as high as 5% yet.

[Alex Bridgeman]

Vesuvio (inc. Capela) 5%

Given your well known preference for mature vintage port, and the relative youth of vesuvio (unless you have a large stash of the 1890'something bottles) I was surprised it was as high as 5% yet.

I was a little surprised, as well. I think it must have come from having taken part in a couple of complete verticals and lots of sample bottles. Plus I tend to have a bottle or two of the latest vintage over Christmas.

[uncle tom]

Average age of VP drunk casually at home:

2005 - 21.5yrs
2006 - 27.0yrs
2007 - 26.6yrs
2008 - 30.8yrs
2009 - 33.9yrs
2010 - 34.4yrs
2011 - 36.1yrs

So far this year, the average age of bottles either drunk or on death row is 42.5yrs

[JacobH]

The most frequently tasted shippers come as no surprise, and are:
Taylor (inc. Vargellas and Terra Feita) 8%
Warre (inc. Cavadinha and LBV) 8%
Graham (inc. Malvedos) 8%
Fonseca (inc. Guimaraens) 7%

Curious that your list ended up almost the same as mine (currently, Taylor, Graham and Fonseca; not sure where Warre fits in). Is this a symptom of UK supermarket discounting policies?

[Alex Bridgeman]

Average age of VP drunk casually at home:

2005 - 21.5yrs
2006 - 27.0yrs
2007 - 26.6yrs
2008 - 30.8yrs
2009 - 33.9yrs
2010 - 34.4yrs
2011 - 36.1yrs

So far this year, the average age of bottles either drunk or on death row is 42.5yrs

Tom's post made me think about the average age of the port I take out of my storage each year. In theory, this should be pretty constant as since 2009 I have had a specific distribution that I aim for and try to manage my consumption to fit that distribution - but I have never checked! So here goes:
2006 - 23.53
2007 - 26.54
2008 - 35.15
2009 - 38.48
2010 - 40.13
2011 - 43.17
2012 to date - 28.14 (expected for the year 39.81)

The target average age for the bottles I pull out in any one year does vary slightly because 10% of my consumption is from my year of birth, so each year I would expect the average age to increase slightly by about 0.1 year. The fact that I'm seeing more volatility than this worries me slightly - especially where the average is higher than expected as this indicates I'm drinking more older wines than will be replaced through the maturing of the other bottles.

I need to think this through a bit more!

[Alex Bridgeman]

I had a few minutes spare this evening between finishing work and meeting my better half - what better way to spend the time than looking at my cellar records?

This time, I decided to look at a different set of statistics. How long have I kept bottles of Port in my cellar?

This is what I found:
11% has been owned by less than 1 year
22% has been owned between 1 and 2 years
29% has been owned between 2 and 3 years
16% has been owned between 3 and 4 years
4% has been owned for between 4 and 5 years
10% has been owned for between 5 and 7½ years
2% has been owned for between 7½ and 10 years
6% has been owned for more than 10 years

No other comment to be made apart from the fact that it's time for me to go and meet my better half. Have a good weekend all.

[DRT]

11% has been owned by less than 1 year
22% has been owned between 1 and 2 years
29% has been owned between 2 and 3 years
16% has been owned between 3 and 4 years
4% has been owned for between 4 and 5 years
10% has been owned for between 5 and 7½ years
2% has been owned for between 7½ and 10 years
6% has been owned for more than 10 years

So!

92% has been purchased since the first tasting at the Crusting Pipe?

{slippery-slope-smilie}

[Alex Bridgeman]

11% has been owned by less than 1 year
22% has been owned between 1 and 2 years
29% has been owned between 2 and 3 years
16% has been owned between 3 and 4 years
4% has been owned for between 4 and 5 years
10% has been owned for between 5 and 7½ years
2% has been owned for between 7½ and 10 years
6% has been owned for more than 10 years

So!

92% has been purchased since the first tasting at the Crusting Pipe?

{slippery-slope-smilie}

Very true. Much of what I had before the first tasting at TCP has since been consumed. And then Tom persuaded me to model my consumption vs cellar age for the next 30 years and the ground slid out from under me.

However, many fewer bottles are being purchased this year compared to the last few years. I am in control. I can decline to buy bargain bottles of port. I can do it, I know I can.

[DRT]

However, many fewer bottles are being purchased this year compared to the last few years. I am in control. I can decline to buy bargain bottles of port. I can do it, I know I can.

Your self-restraint is admirable. But it is a great shame that you must have completely missed out on the recent offers of Dow Bomfim 1992, Quevedo 2007, Croft LBV 2004, Dow's Crusted 2004 and various other LBV bargains

[Alex Bridgeman]

However, many fewer bottles are being purchased this year compared to the last few years. I am in control. I can decline to buy bargain bottles of port. I can do it, I know I can.

Your self-restraint is admirable. But it is a great shame that you must have completely missed out on the recent offers of Dow Bomfim 1992, Quevedo 2007, Croft LBV 2004, Dow's Crusted 2004 and various other LBV bargains

Recently, I have bought no Quevedo 2007, no Croft 2004 LBV and no Dow Crusted. I may buy a bottle or two of LBV in the future but no-one can tell what the future brings.

But I did succumb to the temptation of a case of Bomfim '92. I decided to make an exception for a port that was almost ready for drinking.

[DRT]

Your self-restraint is admirable.

Recently, I have bought no Quevedo 2007, no Croft 2004 LBV and no Dow Crusted

We are not disagreeing.

[Alex Bridgeman]

Typing up some tasting notes, I discovered another area of my spreadsheet which analyses the vintage of the ports I have tasted. My top 5 vintages since 2005 turn out to be:

1970 - 8.7%
1963 - 5.9%
1977 - 5.4%
2007 - 4.8%
1985 - 4.7%

I can only imagine that 2007 is in there from all the sample tasting that I did a couple of years ago.

[Alex Bridgeman]

The end of the year is the time when I post some of the statistics from my cellar spreadsheet so that I have a record for the future.

On 31 December 2012, my cellar comprised 91% Port, 3% Bordeaux, 2% Champagne, 1% Australian, 0.5% of each of Rhone, South Africa, Spain plus small numbers of bottles from England, Sauterne, Hungary, Portugal, Italy, Lebanon, Greece, Chile, US and other French regions.

Of the port, 61% is 21 years of age or older and therefore ready for drinking. 89% of the containers are bottles, 8% are half-bottles, 3% are magnums and I also own 3 double-magnums. 68% of the bottles are vintage port, 25% single quinta vintage port, 6% is LBV, 1% is crusted and there is 1 bottle of colheita, 1 bottle of ruby and 1 bottle of ruby reserve.

The top 5 shippers in my cellar are Vesuvio (15%), Fonseca (8%), Warre (8%), Graham (8%) and Smith Woodhouse (6%). The top 5 vintages are 1963 (10%), 1994 (7%), 1977 (6%), 1985 (6%), 1997 (6%). There are 409 different combinations of shipper and vintage in my cellar.

During 2012 I tasted 433 different bottles of port, which had an average age of 32 years, 101 days. 32 were from 1970, 28 from 1977, 25 from 1963, 21 from 1985 and from 1966. The most frequently tasted port in 2012 was Vesuvio 2010 (8 times).

[RAYC]

Very interesting...

68% of the bottles are vintage port, 25% single quinta vintage port

But how are you classifying this?
eg:
Vesuvio
Noval
Croft
pre-95 Malvedos
1963 Constantinho
Unknown shippers

[Alex Bridgeman]

Very interesting...

68% of the bottles are vintage port, 25% single quinta vintage port

But how are you classifying this?
eg:
Vesuvio
Noval
Croft
pre-95 Malvedos
1963 Constantinho
Unknown shippers

I've tried to be consistent with previous years, although I accept that I have not been internally consistent. However, to answer your questions:
Vesuvio - single quinta
Noval - full vintage
Croft - full vintage
Pre-95 Malvedos - single quinta
(Fonseca-Guimaraens - single quinta)
1963 Constantino - full vintage
Unknown shippers - full vintage, because that's what they have all turned out to be once opened and identified

Any other questions?

[RAYC]

Any other questions?

Is this a challenge to see what your spreadsheet is capable of...?!

If so:

- what percentage of your cellar is SFE and what percentage is TFP?

- are there any vintages or producers represented in your cellar by only 1 bottle? (for the purpose of this calculation, ignoring the possibility that unknowns could turn out to be further bottles from any such vintage or producer)

- for ports of vintage 1970 and earlier, what are the respective proportions of English bottled, Oporto bottled and "Unknown" bottled ports?

- what are the respective proportions of bottles purchased: a) on release b) ex-cellars (but not on release) and c) on the secondary market

- in terms of determining the average age of port drunk to the nearest day, how are you measuring this (e.g.: by assuming [X] September to be a representative average across all producers for when port is "born" each year by fortification of the must to arrest fermentation)?

[Alex Bridgeman]

Any other questions?

Is this a challenge to see what your spreadsheet is capable of...?!

It wasn't intended to be, more of an invitation that if there were other things you wanted to know then you were welcome to ask.

If so:

- what percentage of your cellar is SFE and what percentage is TFP?

56% is from the current portfolio of SFE shippers, and 25% from the current portfolio of TFP shippers.

- are there any vintages or producers represented in your cellar by only 1 bottle? (for the purpose of this calculation, ignoring the possibility that unknowns could turn out to be further bottles from any such vintage or producer)

Yes, lots. There are 17 vintages where I have just one bottle of port from the vintage, and 16 shippers who are represented by just one bottle (treating sub-labels as seperate shippers - eg. Kopke and Kopke Sao Luiz are different). There are two bottles where the bottle is the sole representative of both shipper and vintage.

- for ports of vintage 1970 and earlier, what are the respective proportions of English bottled, Oporto bottled and "Unknown" bottled ports?

This is something that I have not tracked. Where I know the bottler, I have a record but the record is part of the condition report of the bottle and not something that can be seperately analysed.

- what are the respective proportions of bottles purchased: a) on release b) ex-cellars (but not on release) and c) on the secondary market

Assuming that you are referring only to the port, 80% of my bottles have been bought on the secondary market with the remainder being fairly equally split between purchase on release and later purchase ex-cellars.

- in terms of determining the average age of port drunk to the nearest day, how are you measuring this (e.g.: by assuming [X] September to be a representative average across all producers for when port is "born" each year by fortification of the must to arrest fermentation)?

To keep things simple, all port becomes 1 year older on 1 January and remains the same age for the next 365 days. I make no effort to work out the average age of the contents of a bottle to the nearest day based on the day of consumption and an assumed day of harvest or fortification. If a bottle of 1963 port is opened in 2013 then that is 50 year old port, regardless of whether it is opened on 1 January or 31 December. The analysis of the average age of port is the sum of the "age" of all bottles (calculated in the way described above), divided by the number of bottles consumed and the fraction expressed in 365ths.

[PhilW]

The top 5 shippers in my cellar are Vesuvio (15%), Fonseca (8%), Warre (8%), Graham (8%) and Smith Woodhouse (6%). The top 5 vintages are 1963 (10%), 1994 (7%), 1977 (6%), 1985 (6%), 1997 (6%). There are 409 different combinations of shipper and vintage in my cellar.

I'd be curious to know your equivalent percentages for 21yr+ ('drinking ready'); I was surprised not to find Dow in the above (my percentages would start with Warre>Fonseca>Vesuvio>rest, but I expected Dow higher in yours).