Food waste and Scrabble

Summary

More or Less (BBC Radio 4)

Episode Summary: "Food waste and Scrabble"

Date: January 14, 2013
Host: Ruth Alexander

---

Episode Overview

In this episode, Ruth Alexander and the More or Less team tackle two topics:

1. The widely reported claim that up to 50% of all food produced worldwide is wasted—where does this number come from, and is it accurate?
2. An investigation into whether the numerical values assigned to Scrabble tiles are outdated, considering changes to the English language and legal Scrabble words.

---

1. Debunking the “50% of Food Wasted” Statistic

Main Theme:
The episode begins by questioning the validity of the "50% of all food is wasted worldwide" statistic that has been generating headlines and skepticism among listeners.

Key Discussion Points

• Sourcing the Statistic

  • The 50% figure originates from a report by the UK’s Institute of Mechanical Engineers, which itself pulls from a variety of previous reports and studies rather than presenting new research ([00:45]).
    • Wesley Stevenson:

      "The report contains no new figures or calculations. They're all picked from other sources with little regard for how they were worked out." ([01:19])

  • The "up to 50%" number is not precise and is based on broad definitions and older data, sometimes dating back to the 1990s ([01:15], [01:42]).

• Different Definitions of 'Waste'

  • The figure’s credibility depends immensely on what is considered "waste". Definitions can include not just direct losses (e.g., spoiled or discarded food), but also nutrient value losses and even overconsumption ([02:02]):
    • Tuan Tinnemans, Wageningen University:

      "The only way to come up with such a high figure ... is including waste, that is including losses, ... loss of nutrient value ... and probably you also need to include the fact that many people eat too much." ([02:05]) "If you look more narrowly ... the figure ... could be around, I think, maximum 20, 25%." ([02:29])

• Issues with the 30% Statistic

  • The "30%" figure cited is based on more recent UN data but relies on fragmented international datasets and involves numerous assumptions and extrapolations:
    • Wesley Stevenson:

      "Primary data are very scarce and ... most summaries build on few real data points and many assumptions ... the margin of error ... is probably larger than what you would usually see in research projects." ([02:53])

Insightful Moments

• Conclusion on the 50% Figure:
  • The commonly cited 50% food waste statistic is outdated, based on broad and inconsistent definitions, and the true global waste is likely much lower ([03:34]).
    • Ruth Alexander:

      "So do we waste 50% of food worldwide? Case not proven." ([03:34])

---

2. The Mathematics and Modernization of Scrabble Tile Values

Main Theme:
The latter half of the episode explores whether the points values attached to Scrabble tiles are outdated given the evolution of English and tournament word lists.

Key Discussion Points

• Historical Basis for Tile Values

  • Original point assignments by Scrabble inventor Alfred Butts were based on word frequencies in 1930s newspapers and dictionaries ([04:30]).
    • Joshua Lewis, UC San Diego:

      "[Butts] ... looked at the New York Times and dictionaries ... used that to calculate frequency ... and that was used as the basis for the point distribution in Scrabble." ([04:32])

  • Since then, "legal words" in Scrabble have changed, notably with the addition of short, high-value words making some rare letters easier to play ([04:59]).

• Modern Analysis of Scrabble Letter Values

  • Joshua Lewis analyzed modern word lists by:
    • Considering overall frequency, frequency by word length (important because short words influence strategy), and "ease of transitioning" from one letter to another ([05:22]).
    • Joshua Lewis:

      "Letters that were easy to transition to and from were worth less points." ([05:53])

• Findings and Notable Changes

  • Lewis’s alternative scoring system results in only minor tweaks; most values stay the same, but some—like X—would receive a lower score ([06:15]):
    • Joshua Lewis:

      "The letter X ... is worth eight points in the ... retail version of Scrabble ... in my analysis goes down to five points. I think that's very appropriate because X is pretty easy to play." ([06:40])

• Scrabble Community Reaction

  • John Chu, co-president of the North American Scrabble Players Association, believes changing tile values would not be well-received:
    • John Chu:

      "I think there would be catastrophic outrage. ... Changing the colors of the premium squares ... throws people off completely." ([07:01]) "People who've played the game often enough tend to remember ... the Q is worth 10 points, and the Z is worth 10 points, and so on." ([07:19])

  • Chu highlights the distinction between "face value" and "equity value" of tiles, giving the example of the blank tile:

    "The blank in an expert's hands would be worth an average of 30 points to their final score if held onto until they can play it in a bonus play ..." ([07:51])

Notable Quotes & Stories

• John Chu:

  "The clearest example of a tile that is a low face value but has an excellent equity or strategic value is the blank." ([07:51])

• Joshua Lewis:

  "I'm pretty sure I've broken 100." ([08:32])

• John Chu:

  "It was, I think, about 235 points. It was a nine timer. 'Jerkiest.' Even better ... it made my opponent quit the game, quit Scrabble, and nobody's heard from him ever since." ([08:35])

Conclusions

• The values of the Scrabble tiles have barely changed because English frequency remains largely the same, but some tweaks (especially for high-value, now easier-to-use letters) might better reflect current play.
• The traditional values are strongly ingrained in the community and are unlikely to change soon—no plans from Mattel to update the tile values ([08:05]).

---

Timestamps for Key Segments

• 00:28 – Introduction of the food waste debate
• 00:45 – Wesley Stevenson examines the food waste statistics
• 02:02 – Tuan Tinnemans discusses the broad definition of food waste
• 02:53 – Analysis of the 30% food waste figure and its limitations
• 03:34 – Host’s summary: "Case not proven" on food waste claims
• 04:19 – Listener question about Scrabble tile values
• 04:30 – Joshua Lewis on the history and methodology of Scrabble scoring
• 06:15 – Lewis presents findings from updated frequency analysis
• 07:01 – John Chu predicts "catastrophic outrage" if Scrabble values change
• 07:51 – Chu on the real (equity) value of Scrabble tiles
• 08:32–08:50 – Highest scoring Scrabble words and memorable moments

---

Memorable Moments

• John Chu’s humorous anecdote:

  "It made my opponent quit the game, quit Scrabble, and nobody's heard from him ever since." ([08:43])

• The distinction between "face value" and "equity value" in game strategy ([07:51]).

---

Resources:

• Listeners can experiment with Joshua Lewis's alternative Scrabble scoring system using a calculator at the More or Less website.
• Send numerical questions for investigation to [e-mail]

---

Overall Tone:
Informative, slightly playful, and always analytically rigorous—with a keen eye for myths vs. facts and a fondness for numbers that matter.

Transcript

[A]

Thank you for downloading from the BBC. For details of our complete range of podcasts and our terms of use, go to bbcworldservice.com podcasts

[B]

hello and welcome to More or Less on the BBC World Service with me, Ruth Alexander. This week we find out what the consequences would be of changing the value of Scrabble tiles.

[C]

I think there would be catastrophic outrage.

[B]

But before we upset the Scrabble board, lots of you have been in touch asking whether we waste 50% of all our food worldwide. As has been reported, it seems like a big number and a lot of you are suspicious. So is it right?

[D]

No.

[B]

Hi, Wes. Wesley Stevenson's been on the case.

[D]

The figure comes from an unlikely source, the Institute of Mechanical Engineers in the uk. Their report is a complete blizzard of numbers on many different aspects of food waste at different stages of the food processing chain. And the figures are from a variety of sources that use a variety of different methods to reach their conclusions. But if you look at the report, there are actually two figures for overall global food waste. One is up to 50%. So not 50%. Up to 50%, and the other is 30%.

[B]

And so you've discovered that these aren't figures they've worked out themselves.

[D]

No, the report contains no new figures or calculations. They're all picked from other sources with little regard for how they were worked out. For example, the claim we waste up to 50% of the food we produce comes from a UN report published in 2008. But even they the UN took this figure from a book published in 2000 which uses UN data from the 1990s.

[B]

So this bit of news that was being reported so widely isn't really what you'd call new.

[D]

No, it's not very new at all. And the way it was calculated is also in dispute. Tuan Tinnemans is from Wageningen University in the Netherlands and is project director of Fusions, which is a pan European project to tackle food waste.

[E]

If you look at the figure of 50%, then the only way to come up with such a high figure, that means that you're using your widest definition as possible. So that is including waste, that is including losses, that it's including loss of nutrient value. And probably you also need to include the fact that many people eat too much. So overeating could also be an aspect of food waste. If you look more narrowly at the fact of waste and waste is something you don't have a valuable application for anymore, then the figure is far lower. Then it could be around, I think maximum 20, 25%.

[B]

So the conclusion is that the 50% figure is about 20 years old. And even putting that major caveat aside, there are potential problems with the way it was measured. Whereas, what of the 30% figure?

[D]

Well, yes, that figure, which is the other one quoted by the UK's Institute for Mechanical Engineers, that's more recent and comes from a UN study from 20. Now, this study also draws on data collected by many groups and governments from around the world. And when I spoke to one of the authors about its accuracy, he told me that primary data are very scarce and that most summaries build on few real data points and many assumptions and extrapolations have to be made. And he also said that waste is defined in different ways, which means they can't give a quantified margin of error for the 30% figure. But he added that it's probably larger than what you would usually see in research projects.

[C]

Hmm.

[B]

So do we waste 50% of food worldwide? Case not proven. Wes, what's the best score you've ever got in Scrabble?

[D]

77 for tangles.

[B]

I bet you were pleased with that.

[D]

Very pleased.

[B]

But there will be many listening who've done better. The game where players compete to make up words from letters which have numerical values attached to them, is played by millions of people in lots of different languages. But is it flawed? We've had this question put to us by a listener.

[F]

Many new words have come into the language over recent years, but the numerical value assigned to the letters on a Scrabble tile do not appear to have been reviewed to see if their weighting needs to be changed. Does this mean that the results of all recent games of Scrabble have been prone to numerical error?

[B]

This question also struck Joshua Lewis, who's a postdoctoral researcher at the University of California, San Diego. He's been analysing the English language version of Scrabble.

[G]

When Alfred butts in the 30s invented Scrabble, he went and looked at the New York Times and dictionaries and a couple different places where he found examples of the English language in use and used that to calculate the frequency at which different letters appeared. And that was used as the basis for the point distribution in Scrabble. And what's happened in the interim is that the dictionary of legal Plays, or legal words in Scrabble, has changed.

[C]

And.

[G]

And one of the notable additions to me were all of these short words that made it easier to play, for example, z and Q and X. You know, so you have words like qi and za, and so even Though Q and Z are the highest value letters in Scrabble, they're now much easier to play because of these newer tournament legal words.

[B]

So what did you set about doing?

[G]

So in addition to the original analysis that Alfred Butts did of looking at the frequency of letters in the English language, I looked at at the frequency of letters broken out by word length, which is appropriate because two and three letter words and seven and eight letter words are more valuable than the other words in Scrabble. And then I also looked at the ease of transitioning from one letter to another. So for example, with Q, it's very difficult to transition to anything but U. And so letters that were easy to transition to and from were worth less points. And so with these three factors, frequency, frequency by length, and this ease of transitioning, I just took a combination of those statistics and use that to derive value for each letter tile.

[B]

So how does your new value system differ to the current old fashioned Scrabble value system?

[G]

It's actually not hugely different. There's tweaks here and there, but a lot of letters retain their current value. And I think that's a good thing because really, if part of the analysis is the overall frequency of letters, then you would expect things to remain somewhat similar. But there are a couple of notable places for me where things change is the letter X, which is worth eight points in the, you know, the retail version of Scrabble and which in my analysis goes down to five points. I think that's very appropriate because X is pretty easy to play. It's got a lot of two letter words.

[B]

Joshua Lewis, who's been blogging about his new system, but will it catch on? John Chu is co president of the North American Scrabble Players Association. How does he think Scrabble players would react if this new system were adopted?

[C]

I think there would be catastrophic outrage. I think if you look at when seemingly minor design changes are made to the board, such as changing the colors of the premium squares, that throws people off completely. It's possible some people might just keep on playing with the old tile distributions because people who've played the game often enough tend to remember that the Q is worth 10 points and the Z is worth 10 points, and so on. People who are interested in the game already know that the face value of the tile is something that's very different from what they call the equity value. When I teach people how to play Scrabble, first I tell them that they should learn their two letter words. Beyond that, the most useful element of strategy that they can absorb is that they should not pay so much attention to how many points they're scoring on their current play, but they should consider the next three moves in a two player game. The clearest example of a tile that is a low face value but has an excellent equity or strategic value is the blank. The blank in an expert's hands would be worth an average of 30 points to their final score if held onto until they can play it in a bonus play that uses all seven tiles.

[B]

The value as written on a tile is one thing, but its real value could be much higher in an expert player's hands. But back to the question, is catastrophic outrage on the horizon? No. Mattel, which makes Scrabble, told us there is no plan to change the value of Scrabble tiles. But in the battle for the best Scrabble word score, which Wes has already lost, I can tell you who is victorious inventor Joshua Lewis or John Chew from the North American Scrabble Association.

[G]

I'm pretty sure I've broken 100.

[C]

It was, I think about 235 points. It was a nine timer. Jerkiest. Even better though than the score was. It made my opponent quit the game, quit Scrabble, and nobody's heard from him ever since. So that's one of my crowning glories.

[B]

John Chu Victor and if you want to find out what Scrabble tiles are worth under Joshua Lewis alternative system, you can try out our [e-mail] more or less. And if you have a numerical question you'd like us to answer, do email [e-mail]

[A]

There are dozens of different podcasts now available from the BBC, including news, documentaries, science, business, arts and sports. For details of them all go to bbcworldservice.com podcasts.