Why I’m overhauling my credit card review system

I got 243 for my PSLE. 

Maybe to some of you that’s a 3.6 Roentgen kind of score- not great, not terrible. For me, however, it was an unmitigated disaster. It meant that:

1. I would probably end up as a blogger instead of a productive member of society
2. Xiao Ming, my arch nemesis and sometimes 志同道合 friend (but only on 风和日丽 mornings) was 2.88% better than me in life

After all, the numbers didn’t lie. Xiao Ming had a PSLE score of 250, a full seven points more than my ancestor-defiling pittance. 7/243=2.88%. QED. He would now go on to become a Jonny Kim type character, while I would be lucky to survive to adulthood.

And that sort of obsession over fine margins was exactly why in 2021, the MOE overhauled its PSLE grading system. Instead of T-scores with their more than 200 possible permutations, students would now be graded on Achievement Levels, with only 29 possible permutations. No longer would little Timmy be chewed out because Mrs Tay’s son got 10 points more and I need to see her everyday in the lift you know. Granularity was useful, but only to an extent.

I bring this up because it’s an apt analogy for why I’m overhauling my credit card review system…

My current credit card review system

Currently, all my credit card reviews have a score ranging from 1-5 Stars, with half stars also possible. 

While this is a simple, intuitive system, it has one big problem: it cultivates what I call “false precision”. 

I frequently get emails or PMs from readers who ask why one card gets 5 stars while another gets 4 (perhaps that’s the classic Singaporean mindset at play; instead of thinking: 4 stars is great, people ask: where did the last star go?).  I’ve also read debates in the Telegram group where people make the case that Card X should be Y stars instead of Z stars because of A, B and C. 

While I love discussing cards as much as the next miles junkie, I find this somewhat unproductive. The ratings aren’t completely POOMA, in the sense that I’d recommend a 5-Star card much more highly than a 1-Star one, but when you scrutinise them too closely, things start to break down.

For example, I gave the Citi Rewards Card 4 stars and the UOB Preferred Platinum Visa 5 stars, because the former doesn’t pool points and excludes travel from 4 mpd, while the latter pools points and offers an easy route to 4 mpd everywhere

Hold on, someone says. Couldn’t you pair the Citi Rewards Card with Amaze and earn 4 mpd everywhere too (travel aside)? And doesn’t Citi have way more transfer partners than UOB? And doesn’t the UOB Preferred Platinum Visa suffer from S$5 earning blocks, UOB$ merchants and UOB’s other dirty games?

Yes, yes, and yes. Those are all completely valid points, and that’s just the tip of the iceberg. We could have lively debates as to why the HSBC Revolution warrants 4.5 stars when it adopts a more restrictive “whitelist approach” compared the “blacklist approach” of the 4 star Citi Rewards, or why the DBS Woman’s World Card gets “only” 4.5 stars when it’s a simple route to 8,000 miles each month.

But this entire exercise is pointless. Why quibble over fine differences when at the end of the day, there’s nothing stopping you from getting both? This isn’t meant to be a beauty pageant where a single winner gets crowned, after all. Quite the opposite, it’s more of a harem situation where you decide which cards make the cut (and which get the cut, a la Anne and Catherine).

So the problem with the current system is that it suggests a kind of granularity that just isn’t there, and if a missing star is what stops you from getting a Citi Rewards Card, there’s something very wrong.

My new rating system

When I think about credit cards, I don’t think about 3, 4 and 5 stars. Instead, my thought process is much simpler:

• Is this a card I’d insist people get?
• Is this a card I could see being used as part of a sound miles strategy, but perhaps not absolutely essential? 
• Is this a card I’d warn people to stay away from?

With that in mind, all we really need are three categories. 

Take It are cards that I consider an essential part of my miles collection strategy. These see the heaviest use, and if any one of them got nerfed, there would be an immediate and noticeable decline in the number of miles I earn.

Some cards I’d consider part of this category:

• Citi Rewards
• DBS Woman’s World Card
• UOB Preferred Platinum Visa

Leave It are cards that I find underwhelming because of some deal-breaking flaw, or the fact you can find an alternative which beats it in almost every way.

Some cards I’d consider part of this category:

• BOC Elite Miles Card
• StanChart Rewards+
• UOB Visa Infinite Metal Card

Take It or Leave It are cards that I may have applied for during a sign-up bonus or some other limited-time offer, but don’t see regular use outside of that. They remain on standby until another offer comes along, like the periodic bonuses offered to UOB PRVI Miles Cardholders. 

Some cards I’d consider part of this category:

• Citi PremierMiles Card
• DBS Altitude Card
• UOB PRVI Miles Card

Perhaps the main criticism of this approach is that “Take It Or Leave It” could become a catch-all pile. Most general spending cards will end up here (unless they have some excellent limited-time promotion, like the OCBC 90°N Mastercard and its launch offer of an uncapped 4 mpd on overseas spend, SIA & Scoot tickets), as will the $120K cards + AMEX Platinum Charge. 

But maybe that’s the point: I don’t consider any of these to be essential. General spending cards should be your card of last resort, and while I make heavy use of my AMEX Platinum Charge’s benefits, I can’t definitively tell someone they should get this (or a $120K card) given the high annual fees, and how subjective the valuation of perks can be. 

Ultimately, in the absence of a detailed rubric to dictate how stars are awarded (e.g. +0.5 stars for points pooling, -0.5 stars for S$5 earning blocks, -30 stars for PCK crooning in your ear all day), the current system strikes me as too arbitrary. There’s certainly a difference between a 5-star and 1-star card, though the nuance between a 3-4 star card becomes less meaningful. 

I imagine not everyone will be happy with the change, but put it this way: if you used the star system as a rough guide of what to get or avoid, then the new rubric still does the job. On the other hand, if you believed that a 3.5-star card was indisputably better than a 3-star card, then the star system was steering you wrong anyway.

And to the extent that a “Take It or Leave It” label encourages people to read the full review, I only see that as a good thing. Too much nuance is lost in a single headline figure. 

Conclusion

I’ll be introducing a new way of reviewing credit cards that does away with the 1-5 stars in favour of a less granular approach. It’s my hope that this will help people think about which cards they should definitely (and definitely not) get, while avoiding the pitfalls of seeking out a single “best card”. 

It won’t happen all at once — the existing scores will remain as is until I get around to updating each of the reviews for 2023 — but that’s the way I see myself scoring cards going forward.