Marriott has now jettisoned the last remnants of their old award charts by removing top-end award price caps. Stephen collected data before and after the cap was removed and discovered that the results weren’t nearly as bad as feared. Yes, some hotels are sometimes much more expensive, but they’re often less expensive as well. See: Marriott’s 2023 “Devaluation” – Much Better Than Expected.
I took a different approach to analyze the changes. I used the same methodology as I’ve used in the past to calculate the value of hotel points to see if Marriott’s average point value has changed. The quick answer: not really. Point values dipped a tiny bit, but not enough to change our Reasonable Redemption Value. The value of Marriott points, based on our Reasonable Redemption Value methodology, remains at 0.8 cents per point.
When collecting points and miles, it’s always a good idea to have a general idea of what points are worth. Let’s say, for example, that you have the opportunity to either earn 1,000 Hyatt points or 2,000 Marriott points. Which should you go for? If you don’t know what the points are worth, you’d likely go for the Marriott points. But, in my analyses I’ve found Hyatt points to be worth more than twice as much as Marriott points. Therefore, on average, 1,000 Hyatt points are worth more than 2,000 Marriott points. In this post, you’ll find my best current estimate of the value of Marriott points. To see our estimates across many programs, see this post: Reasonable Redemption Values (RRVs).
In order to determine the value of Marriott points, I collected real-world cash prices and point prices. As I’ve done previously, I examined a number of major hotel markets in the U.S.: Chicago, Denver, Los Angeles, Hawaii, Las Vegas, Miami Beach, New York City, Orlando, and Seattle. Within each market, I identified the first three search results with a guest rating of 4 or better, and I recorded both cash and award prices for three dates each: a weekday, a weekend, and a holiday.
- Why U.S. only? U.S. consumers are known to spend most of their points and miles on domestic travel. Since the majority of this blog’s audience resides in the U.S. we opted for a U.S. centric view of point values.
- Why Guest Ratings? The goal wasn’t to find the 3 best Marriott hotels in each market. Instead, the goal was to find the 3 Marriott bookable hotels that are very popular. Which ones are people really likely to book? By using a combination of Marriott’s default sorting and by picking only well rated hotels, I think it’s reasonable to assume that many members would pick these hotels.
- Which paid rates were selected? I always picked the best fully refundable paid rate shown on Marriott’s website, but without applying any discounts like AAA, military, government, etc. For this analysis, I usually selected Marriott’s Member Flexible Rate.
- Which specific dates did I use?
- Weekday: Wednesday Nov 8, 2023
- Weekend: Saturday Nov 11, 2023
- Holiday: Thursday Nov 23, 2023 (Thanksgiving)
Starting 12/5/2022 with my Hyatt analysis, I began calculating Cents Per Point (CPP) using new formula. In all past hotel CPP calculations, I didn’t account for taxes and fees or points earned on paid stays. I figured that those things basically cancelled each other out. Going forward, I am including these factors, as well as resort fees, in the CPP calculation. The calculation is based on the following terms:
- Base Cash Rate: This is the hotel room rate before taxes and fees.
- Total Cash Rate: This is the total amount, including taxes and fees, that would be paid if booking a hotel’s cash rate.
- Resort Fee: This is a fee that is imposed by many hotels above and beyond any required taxes. This goes by different names at different hotels: Resort fee, Destination charge, Founders fee, etc.
- Points Per Dollar Earned: The number of points per dollar earned by non-elite members on paid stays. For example, Hyatt members earn 5 points per dollar, Hilton, IHG, and Marriott members earn 10 points per dollar (at most hotels), etc.
- Points Earned on Cash Rate: This is the number of points you would earn if you paid the cash rate. The calculation for this is: (Base Cash Rate) x (Points Per Dollar Earned). For this calculation, our default approach is to assume that the traveler does not have elite status (elite members earn more points per dollar).
- Point Price: The number of points required to book a night at the hotel
- Cents Per Point (CPP): This is the value you get per point when using your points instead of cash to pay for a stay.
Hotel Programs that Waive Resort Fees on Award Stays
Hilton, Hyatt, and Wyndham waive resort fees when you book stays using points or free night certificates. For these chains, the resort fee does not have to be considered separately from the Total Cash Rate (which includes the resort fee). So, the CPP calculation is as follows:
CPP = Total Cash Rate ÷ [Point Price + Points Earned on Cash Rate]
Hotel Programs that Charge Resort Fees on Award Stays
IHG, Marriott, and many other hotel programs impose resort fees on award stays. For these chains, the resort fee must be specifically taken into account in the CPP calculation. We do that by subtracting it out of the Total Cash Rate. The CPP calculation is as follows:
CPP = [Total Cash Rate – Resort Fee] ÷ [Point Price + Points Earned on Cash Rate]
|Point Value (Median)||0.76||0.81||0.75||0.66 cents||0.63||0.62|
|Point Value (Mean)||0.80||0.85||0.81||0.73 cents||0.66||0.67|
|Cash Price (Median)||$419||$408||$303||$284||$246||$246|
|Cash Price (Mean)||$462||$464||$392||$305||$303||$282|
|Award Price (Median)||51,500||48,500||42,000||50,000||42,500||40,000|
|Award Price (Mean)||53,244||48,444||46,761||44,203||46,212||44,802|
|Minimum Point Value||0.38||0.43||0.31||0.34 cents||0.26||0.30|
|Maximum Point Value||1.65||1.78||1.71||1.89 cents||1.38||1.67|
* Analyses prior to 12/5/22 relied on pre-tax numbers and didn’t account for points earned on paid stays. For this reason, I grayed out the numbers that can’t be fairly compared to the most recent numbers.
The median observed point value for the latest analysis was 0.76 cents per point. This means that half of the observed results offered equal or better point value and half offered equal or worse value. Another way to think about it is that without trying to cherry pick good awards, you have a 50/50 chance of getting 0.76 cents or better value from your Marriott points when booking free night awards.
In my last data collection in December 2022, the median cents per point value was 0.81. Since both the previous median (0.81) and the new (0.76) round to 0.8 cents per point, we’ll keep our Reasonable Redemption Value at 0.8 cents per point.
Pick your own point value
|50th Percentile (Median)||0.76||0.81||0.75||0.66 cents||0.63||0.62|
|60th Percentile||0.81||0.87||0.80||0.73 cents||0.66||0.66|
|70th Percentile||0.88||0.93||0.91||0.80 cents||0.73||0.77|
|80th Percentile||0.95||0.99||1.08||0.85 cents||0.83||0.81|
|90th Percentile||1.15||1.15||1.21||1.03||0.98 cents||0.96|
When we publish Reasonable Redemption Values of points (RRVs), we conservatively pick the middle value, or the 50th percentile. The idea is that just by randomly picking hotels to use your points, you have a 50/50 chance of getting this value or better. But what if you cherry-pick awards? Many people prefer to hold onto their points until they find good value uses for them. If that’s you, then you may want to use the table above to pick your own point value. For example, if you think that you’ll hold out for the best 10% value awards, then pick the 90th percentile. If you cherry-pick a bit, but not that much, you might want to use the 70th percentile (for example). I’m guessing that most cherry-pickers will land around the 80th percentile: 0.95 cents per point.
To me, this analysis shows that those who cherry-pick good value awards can count on getting around 0.95 cents per point value, or better.
Reasonable Redemption Value: 0.8 Cents Per Point
Our Reasonable Redemption Value (RRV) for Marriott points was previously set to 0.8 cents per point and the new analysis confirms the same. RRV’s are intended to be the point at which it is reasonable to get that much value or better for your points. Therefore, I believe that the median observed value for is a good choice for our RRV.
- Reasonable Redemption Value for Marriott: 0.8 cents per point
- Reasonable Redemption Value for those who cherry pick awards: 0.95 cents per point
Overvaluing vs. Undervaluing Points
There is no perfect way to estimate the value of points. Decisions we made here in some ways overvalue points and in some ways undervalue points. The hope is that these things roughly offset each other…
Factors that cause us to undervalue points
- With hotel programs that offer 4th Night Free Awards (IHG, with some credit cards), or 5th Night Free Awards (Hilton & Marriott), or award discounts (Wyndham), we do not consider the point savings in our analyses.
- With hotel programs that offer free parking on award stays to top-tier elites (Hyatt), we do not factor this in.
Factors that cause us to overvalue points
- We do not use discount rates (other than member rates) in our analyses. In real-life, many people book hotels cheaper (and sometimes far cheaper) by using AAA rates, government & military rates, senior rates, etc.
- We do not use hotel promotional rates. Often, individual hotels have deals such as “Stay 2 Nights, Get 1 Night Free” which can greatly reduce the cost of a stay.
- We do not use prepaid rates in our analyses. Sometimes these rates are significantly lower than refundable rates.
- We do not factor in rebates which can be earned from booking hotels through shopping portals.
- We do not factor in extra points earned on paid stays for those with elite status.
- We do not factor in rewards earned from credit card spend at hotels.
- We do not factor in hotel loyalty program promotions: Most promotions, but not all, only offer incentives for paid stays. We often see promos offering bonus points, double or triple points, free night awards, etc.
- With hotel programs that waive resort fees for top tier elites on paid stays (e.g. Hyatt), we do not factor this in.
Based on the latest analysis, I’ve kept our Marriott RRV the same as before: 0.8 cents per point. The idea is that you have equal chance of getting that much value or more from your award stays. That doesn’t tell the whole story. The latest percentile results show that it’s reasonable (80th percentile) to expect to get 0.95 cents per point or better if you cherry-pick good-value awards.
For a complete list of Reasonable Redemption Values (and links to posts like this one), see: Reasonable Redemption Values (RRVs).
In your cherry picking table you have the most recent measurement date as 04/07/22, should be 23?
Yep, thanks. Fixed.
I just spent 17 nights in Qatar during the World Cup and got incredible value from my Marriott points.
I believe the value abroad is far better, particularly compared with the US West Coast.
Should be called MARRIOTT PESOS, IMHO.
Pretty close to my valuation. I tend to cherry pick and came to about a 0.9 cent to 1 cent value for Marriott points.
Interestingly, the closer to 1 cent you get for Marriott, things like the Chase UR transfer bonus can make sense in certain situations. I transferred over a good chunk of points for a Hawaii redemption at well ove 1 cent per point.
I know a lot of the hub bub has been addressed to the higher priced hotels. What is the impact on travelers to more rural/suburban areas? Many self employed people try to trim their costs using FFI, Courtyard, or other monikers. They may travel to second or 3rd rate cities and may be restricted in expenses by their employers.
Some of the discounts didn’t last long. My reservation went down by 500 points the first day, then up by 3000 the second day (for now) indeed!
Booked Maldives St. Regis before change. 400k. Post change same dates 685k
I locked in five nights in June at the Westin Hapuna Beach resort before the devaluation. Total before was 210,000 points for five nights. Today it is 298,000 for the same five nights. Ouch! Thank goodness I locked in the redemption price. Once the “top off” is possible, I plan to top off my five 35K free night certificates to apply them to the reservation and get 175,000 points back into my account. I talked to a Marriott rep before locking in the reservation and was told this was completely possible. Hope it works. Otherwise, I’ll be holding five worthless 35K certificates, as they have to be used before the end of June.
Between yesterday and today some hotels went up in points. I was looking at Washington DC and Ocean city MD.
To all the folks out there cancelling Marriott reservations and rebooking for fewer points – there is a much easier way. You can edit your existing stay in two clicks and the difference in points is redeposited into your account immediately – and the reservation number stays the same, which is handy if you’ve reached out to the hotel for a rollaway or whatever. This method also saves you the time of searching just to see if you’ll save points because you’ll see the new points requirement immediately for the hotel, room type, and dates you’ve already booked.
Just click on “view/modify” on the given reservation under My Trips then “Edit room” and you can see what the new points requirement is. Click “Update” and you should be just about all set. I saved over 100k points yesterday on 6 reservations thanks to Marriott’s new points system and about 5 minutes of time.
Note – on one of my reservations, editing the room seemed to work but did not, the points charged did not change to the lower amount for some reason. I did actually have to cancel and rebook that one, but all of the others I was able to adjust the points needed downwards just by editing the room.
Thanks for posting this, I went and checked my reservations and 6 out of 7 were actually cheaper than before and this made it easy to get the difference in points back to my account.
I am totally agreed your analysis. So far tt looks good for the dynamic pricing. To be honest, I just updated my previous booking and saved myself 17,000 points redemption after I cancelled and rebooked my stays in Canada.
My Protea Kruger Gate hotel went down by 1,000 Points! Woohoo big win not, but I did rebook it and cancel the other booking though.
I’m in the process of booking a hotel for April 14 and out the 18th at the New Orleans Warehouse District and the price points increase for Saturday night form 31,00 to 33,000. its not a lot but still that shows the dynamic pricing in place
Sure, it’s not bad if you use points instead of cash at mediocre destinations. However, I use points to stay at properties I would never pay cash rates at. These are the properties that have been most devalued. For example, I just stayed at the Alila Napa. I’d never stay there at $1600 a night, just ridiculous. But points? Sure, why not.
This deval is huge, it takes the fun out of earning points. It’s a cash rebate program now. All aspirational properties are now out of reach. In addition, the FN certs are basically unusable at anywhere decent. If they are going to deval all the good FN cert options, then I’m not going to be carrying their cards for a FN cert that will need to be topped up basically anywhere I’d like to stay. Bottom line: your averaging across domestic locations methodology is a poor indicator of the program devaluation because many people here tend to use points for aspirational stays. I’m done with Marriott after I burn these last FN certs.
I also agree with you that who would pay 1000$+ a night for Alila Napa. I stayed there twice now, it’s a nice property but def not worth 1000$+ cash. Maybe for people that just doesn’t care about spending money?
Glad I redeemed two 50k certificates for the JW in Marco Island back in January for a June stay.
3 nights in Turks and Caicos 255,000 yesterday, 325,000 today.
Not so good 🙁
Can we “top off” yet?
Not until “late April”
I collect hotel points mainly for use abroad. I’m not sure FM readers “spend most of their points and miles on domestic travel.”
Are you saying FM gets the most traffic when you talk about using points for domestic travel, such as a Spirit flight to the Omaha overwater bungalows?
“surprisingly good”? Maybe “less bad” is more fitting.
Why do you say that?
Great analysis, Greg.
However, I am not as hopeful as you when you say: “And once Marriott enables the ability to add up to 15K points to our certs, I expect that the value of these certs will go up quite a bit.”
I bet that by the end of the year Marriott will dot what Hyatt has just done and downgrade these annual credit card certs. For me in California the Hyatt devaluation is bad and Marriott is already a problem finding somewhere decent for the basic cert from the credit card.
Nick& Greg, Any additional data points for this article? I’ve called Marriott and the folks that I get don’t ave any clue about this program.
I think your methodology substantially overvalues certificates. They expire worthless if not used withing a short time, they are for exactly one night, no flexibility, and especially with the bigger certificates, it’s rare to have a hotel worth spending that much cash on, at a time and place where the certificate could be used instead.
In practice, my cat 1-4 certificates end up getting used at airport hotels in lieu of cash rates of $120-$150 including taxes. I haven’t had any larger certs from Marriott., but my “unlimited” Hilton certificates from the Aspire card ended up used in ways I only considered worth $200-$250 each (to be fair, if I had planned for the pandemic, I could have gotten more value).
I agree. In this post I used data to estimate the median value of how much the certs are worth towards hotel redemptions at the top point value plus one or two lower point values (e.g. I looked at 85K certs being used for 85K and 70K and 60K properties). I think that the values I came up with are reasonably accurate for that purpose: If you use the certs in that way, you can expect to get about that much value.
The redemption value is a very different metric than “what are the certs worth to you?” When estimating how much they are worth to you, it’s important to include factors like: what is the chance that they’ll expire and go unused at all? what is the chance that you’ll end up using them on a cheap property? How much would you buy these certs for if they were for sale? etc.
I’m still noodling about how best to come up with values for the blog that can be used as estimates of the overall value of these certificates. My best current thinking is to use the RRV and apply a fudge factor that represents how much less valuable certs are compared to points. For example, we might decide that certs are worth 2/3rds of their point value, like this:
Those, I think, are reasonable numbers that can be used as estimates of “value” where value means “I would be equally happy to have the cert or that amount of money”. Or, yeah, I would consider buying a cert, if it was for sale, for that amount of money even without definite plans for how I’d use it.
Great way to force clarity on cert value by asking how much you would pay for a cert that you have no definite plans for. Key is assuming you have to buy NOW to get the price. That forces you to truly buy with no plans in mind
Interesting, this basically means that when you factor in the increased annual fee, the Brilliant 50k cert isn’t really more valuable than the Boundless/Marriott AMEX (ex-SPG, not available for new customers) 35k cert- you net $55 in value on a $95 AF vs $60 in value on a $150 net AF (minus the $300 in Marriott charges you get comped).
It feels like an argument to dump the Brilliant for those cards (unless you really value Marriott Gold over Silver, which is the only major thing you get from the Brilliant). Perhaps this is why they were surveying people about a boost in Brilliant’s AF + boosting the certificate value…
Great analysis Greg! I wish you would include a few international markets. First, I think the audience of this blog is not “average” and most probably do use them internationally. Second, you include 5 different percentiles to show the potential of cherry-picking and yet leave some of the best cherry-picking redemptions out. I’d be interested to see the 80th percentile considering those markets.
One odd thing I noticed — the value of the 85k certificates is higher than the RRV of just having 85,000 Marriott points. Same for some of the other certificates values. I suppose you could say that implies doing some cherry-picking with the certificates, but still odd nonetheless.
Yeah, I was thinking about that this morning. I think it suggests that 70K and 85K properties (in my dataset) tend towards higher point value than average. That makes sense to me because we often see better value for points when used for top-tier hotels since real world prices aren’t capped, but award night prices are capped. That said, it does lead to a problem with using these numbers to make decisions. If you went entirely by my numbers, you would value 85K points and certs as follows:
And then you would presumably make the mistake of thinking that a signup bonus with an 85 cert is more valuable than a bonus that includes 85K extra points. In reality, the reverse is true: any sane person would far prefer 85K points over an 85K certificate. For the purpose of our first year value calculations, we’re using a much lower number for the certificate so as not to lead people into making this error.
Just curious- why don’t you include taxes and fees since those would be paid w cash but not points? It seems to not include the full price always under values the points! I understand not including resort fee since isn’t widespread, there are many hotels (even in markets w resort fees) that don’t have them and some elite members may be exempt. However it frankly seems sloppy to not include basic taxes and fees everyone would pay on the cash price especially when it is so easy to determine that.
I tend to only use points when I get more than my personal value (I have Marriott at .7 cent) but make sure to include the taxes when determining if to use points or pay cash for a stay.
Did you read the section titled “Overvaluing vs. Undervaluing Points”? That’s where I tried to explain the decision. Is there something I can do to make that section more clear?
It would involve significantly more work, and might be worthy of a post on its own, but I would like to see those factors quantified with a few real-world numerical examples (perhaps across different chains). I see the point that not including taxes (and resort fees, etc.) might overvalue points because of the countervailing factors you list, but it feels a bit hand-wavy to just say “all these factors average to zero or close enough” without actually running some numbers.
(Maybe you’ve done such analysis in the past and I just missed it.)
One factor with the high value certs is that who just wants one night? I find this with the Hilton free night certs. I never just want one night and when I consider a thee night booking — one cert and 190,000 points for example — I end up preferring using different points at a different hotel. I recently used a Hilton cert for a 60k hotel for just this reason, for about $400 in value. I could have gotten $1,000 in value, but who wants one night at the Grand Wailea? Bigger picture, free night certs stress me out. They cause me to make suboptimal decisions. Points are way better, even fewer points.
Marriott not honoring elite benefits makes certs of zero value to me
Just dropped 400K points for 5 nights at Al Maha, where rates are $2300 per night. Lot of points but good value. 150K at LeMeridian for overwater where rooms are nearly $700 per night. Lastly, 240K points in Hawaii where rooms are about $600/night. Not best value in Hawaii, but OK.
Ouch! Marriott takes another hit!
Greg: I recently redeemed 70,000 Bonvoy point per night at the Ritz Carlton Key Biscayne for this upcoming Memorial Day Weekend. Room prices exceed $1,300/night which would equal over 1.8 cents per point. I’ll take deals like this everyday… All the best, RM
Just wondering about the deeper reasons behind this. Since this and most of the other hotel brands you (@Greg The Frequent Miler) featured have devalued their currencies, I wonder if there might be another reason involved than just hotel “greed”. Do you think it’s possible that cash prices may have come down due to the pandemic, and therefore inadvertently caused a points devaluation? All we’d need for that to happen would be less or no drop in points/award prices when there is a drop in cash prices.
Well, Hyatt and Wyndham RRVs went up. I’m not sure that anything has really changed other than the methodology for figuring out the RRV. I don’t think the drops are due to lower pandemic prices. In my experience, most hotels are pricing higher, if anything. There’s a boom in domestic travel I believe.
I have two of the 25,000 point certificates and they seem pretty worthless to me.
If you ever go visit Oklahoma
I redeemed 210k points this fall for a 5-nt stay at the Mauna Kea BI. Well over .02/point value. If I earned points for nothing other than St Regis & RC redemptions, I would have similar results. I usually pay cash for Towneplace Suites, etc & earmark points for a long vacation with great redemption value.
I have been getting around .75 cents redeeming at a Westin that is now 40,000 points per night, and I would actually pay cash to stay there so it’s real (so around 70th percentile from your analysis).
About 70% of the time I transfer Marriott points to airlines. I want to get one more Emirates First Class trip using Japan Airline miles so I will be transferring points soon. The other 30% are cherry picked stays on points with very good value. I think there is good overall value.
You should do this again! Please, I beg of you!
[…] restaurant within 30 days, and leave a review. That’s not a huge bonus given that Greg recently determined Marriott points to be worth 0.7c each (about $7), but it’s better than […]
On a call with a Marriott customer center rep, he offered me “as a one-time courtesy” to exchange my 50k free night certificate for 50k points, which I accepted. The exchange happened the next day.
Wow! How recent was this? I’d take that anyday, and my wife and I are each sitting on a 50k certificate.
[…] afternoon, Greg published a post pegging the current value of Marriott points at 0.7c each (See: What are Marriott points worth?). If you agree with his valuation, this is like 4.2% back on grocery spend, which can be a solid […]
“I haven’t seen a drop in cash prices for resorts in the United States. At least, not yet.”
It’s coming. Hotel prices and airfare are subject to the same supply and demand economics like anything else in the marketplace. IF travel demand continues to be severely curtailed, prices will have to fall to try and stimulate demand. That means they will have to lower the cash price AND the number of points required to travel somewhere to increase the weak demand. Therefore, the points you currently have in your account will have to increase in value because of all the “deals” that will eventually be offered. Eventually supply and demand will balance and then a new re-evaluation of point value can be determined. One also has to keep in mind the effect of all the millions of job losses that will also probably depress travel demand. First, people need their jobs back and their businesses to re-open. Then they want their healthcare. Then they have to pay off some of their debt they accumulated while unemployed. Then, maybe, they’ll start thinking about taking a vacation and how much it costs to go on one. I don’t think this will be a V-shaped recovery process for the travel industry, but a long, slow ordeal. Just my opinions, nothing more.
Hope u got a haircut before video.
How’s the son doing on travel&Points? I have a college person across street laughs @ points.
Is this a trick question. I would say, not much.
Marriott just swept the Hotel category at the Freddie Awards!
I’ve thought about this myself and I would argue that the value for Marriott points has a floor of 0.7cent/point simply based on the value of some of their airline transfer partners.
For example, with Alaska Air, 60000 points converts to 25000 miles which is .416 miles/1 point. Multiply by the rough valuation of Alaska miles on sale being around 1.8cents/mile and you get .75cents/point. Since Marriott points can be passively gained from stays/CC spend, I would much prefer to convert 60k points than spend $450 to get those 25000 miles. I’d expect similar scenarios being true with other miles.
The irony is that, with all the devaluations from Bonvoy, the reward for loyalty to their chain isn’t even tied to staying at their own hotels. I’ve found it hard to get good consistent .7/point value on my Marriott bookings without going through a significant amount of work. While .7/point is the average value, I can’t expect to pick any Marriott at my travel destination and expect to get that value. On the other hand, I can transfer to miles at any time that suits me and even with a mile devaluation, premium class cabin redemption and redemption sales should still get me more than 1.8 cents/mile.
Great points. Thanks!
You really can’t use the sale price of AS miles in your calculation though, unless you are saying you use the points exclusively to top up accounts where you would otherwise purchase the miles (and that you always do it in 60,000 Marriott point increments). Otherwise, you could just as easily say that Marriott points are worth ~0.85cents since that is the sale price they have sometimes been sold at.
Greg’s RRV post values AS miles at 1.3cents. So to be consistent on that basis, the floor for Marriott point value on that basis would be 1.3 x 25000 / 60000 = 0.54cents. (Obviously could be different for you based on your own valuation of AS miles!)
(I actually also find that most of the value from Marriott points come from airline transfer partners. I have actually never used a single Marriott (or SPG before the merger) point at a hotel.)